Over the next two posts, I’m going to describe the project lifecycle after the sale is complete.
As case studies, I’m gonna use two of my recently-launched projects: one a major rebrand / web redesign / dev project for several sister companies, and the other a much smaller design & dev project that resulted in an ‘internal prototype’ for a publishing group.
This first post tackles the bigger project and illustrates all the ins & outs / phases / considerations a project might have. I’ll walk through a very comprehensive framework for organizing a high-complexity, high-budget job.
The next post will describe the smaller project, and distill the comprehensive framework to the key pieces needed to do quality work within a limited budget and timeframe.
After these posts, I’ll give more detail on:
—hiring developers and designers
—PM best practices
—pricing
—some of my key organizational techniques.
These case studies should provide some helpful context for the points I’ll make in those later posts.
Alright. Down to business.
PART 4: The Complete Project Lifecycle
![[Image: UIQBSrL.jpg]](http://i.imgur.com/UIQBSrL.jpg)
4.1 From Generalist to Specialist & Back
If you follow my career trajectory and start from the bottom, you’ll begin by handling every aspect of a project: design, development, content insertion, launch, and sometimes ongoing maintenance. You might throw in some copywriting or set up the client with a professional photographer.
You need those first clients' business, so you give them whatever services they ask.
After awhile, you’ll probably find it more profitable to specialize in one or two phases of a project.
For example, I soon began signing clients, working with them to create a written design spec, hiring a design agency to take 95% control of the project during wireframing and design, and then taking the wheel again for development.
Development was what I knew and where I earned my money, so specialized in it.
During that time, I was getting more complex projects, but I didn’t have any connections to hire designers directly. Even if I could’ve hired them, I didn’t have the experience to supervise their work. Outsourcing to a reputable agency was my best option.
The drawbacks to that were A) my company wasn’t making money off design, B) because of the agency’s markup, the clients and I were paying $150 / hour for $50 / hour junior design work, and C) I lost control during design.
Which sucked. So I reached out to design folks, tried some out, found a freelance Creative Director, and assembled a team of freelance designers directly responsible to him & me. Increasingly I cut out the design agency on projects I personally pursued / landed.
As luck had it, my Creative Director had a reputation for being an excellent web strategist (“design is strategy” is a common refrain in this industry). He also had a working relationship with a well-regarded guy who freelanced on projects as a “Head of Strategy.”
With their help, I started landing a whole new kind of client: Instead of clients who came to me with detailed specs and long lists of demands, they would pay us to A) tell them what to do and B) execute our recommendations.
The transition was akin to going from a carpenter, who works off someone else’s blueprint, to an architect who has the vision and authority to chart the creative course. Excuse the shitty analogy.
With that shift, my company was once again overseeing the entire course of projects: initial research & strategy (aka “Discovery”), design spec, design, development & launch.
4.2 Project Types
I explain all this to show that what I uniformly label as “projects” actually come in many different types. On two simultaneous engagements, my company might handle completely different kinds of work.
On some projects, we only do dev. Others, we do only design.
Some, we've done wireframes (initial outlines of designs), let another firm do final designs, and then handled dev.
Once, I ran Discovery and formulated a design & tech spec, which the client turned into an RFP and opened for bidding. I then lost the bid process and had nothing to do with rest of the project.
When I say “project lifecycle,” I’m referring to the full extent of my own shop’s involvement in the project.
As you can see, that’s sometimes only one phase, and sometimes it’s every last aspect.
4.3 The Client
The job I’m outlining here was the full-service sort. A family of four outsourced HR firms (I call them Companies X1, X2, X3, and X4 below) had retained a major NYC agency to conduct a Discovery round and propose a plan for complete rebrand and web redesign.
Assuming the plan was half decent, the agency would’ve transitioned directly into the $150k rebrand/redesign/development project.
In fact, the big agency treated the HR firms like small potatoes and gave them a boilerplate plan. HR firms balk; big agency gets fired; LouieG & Friends step in to the save the day.![[Image: lol.gif]](https://rooshvforum.network/images/smilies/new/lol.gif)
The HR firms wanted my company to start from scratch. We’d run a new Discovery phase, create our own plans for branding and web redesign, and then get all the branding / design / dev work, provided our plans didn’t blow.
(No need to worry about that big agency. They picked up where they left off: Handling branding and websites for companies with multibillion dollar market caps.)
4.4 Discovery Phase
The purpose of a Discovery Phase is to A) identify specific project goals, B) create a list of content-to-create / features-to-develop that satisfies those goals, and then C) bend the content & features until they fit within the client’s budget and (when applicable) technical capacities.
The deliverable is usually some combination of:
—Website plan: A written doc including content overview, page templates, feature set & functionality, and information architecture map
—Design overview: Describing the style of web design and likely design elements; look for words like “playful,” “modern / clean,” “spacious,” “savvy,” “humorous,” “aspirational."
—Branding plan (when branding is in scope): Similar language and intent to the design overview. Also includes a list of branding assets to make: logo, brand guidelines, letterhead, business cards, email signature.
—Tech specs: CMS / hardcoded? Stripe / PayPal / Authorize.Net? Responsive web design? Retina ready graphics? If there are ambiguous functions or workflows, this is an opportunity to spell out how they’ll work, step-by-step.
Discovery isn’t a necessary part of every project. If the client comes to me knowing what they want in every detail, I skip it.
If they’re expecting me to generate ideas, I'll insist on a Discovery phase. Finding/refining ideas and getting them approved by the client is a time-intensive process that absolutely should be built into a project’s pricing and timeline.
With regard to the specific project I’m discussing—I won’t say much about contracts here, but as a note on process, we began with a $6k deal for 60 hours of discovery work. $4k on signature, $1k on delivery of rough draft plans and final draft plans.
No guarantee we’d get a dime beyond that, although we & the client were hoping the relationship worked out.
So this particular Discovery phase went as follows:
Meeting 1: Kickoff
Actually, five kickoff meetings: one with the stakeholders from all four sister companies, then breakout sessions with the individual companies.
A well-run kickoff meeting includes a lot of ‘handwaving’—business jargon for shit you do simply for appearances, to reassure the client you’re doing job. We explained the Discovery deliverables, outlined our process & the meetings we’d conduct, and spoke about our belief in the importance of doing good work and really listening to clients.
The client, for their part, bitched about their poor treatment from the previous agency, complained about their self-evidently terrible websites, and told us how they believe in doing good work and really listening to clients.
Signal-to-noise was 1:100. Kickoff meetings are like that.
The important thing was that, somewhere amidst all the platitudes, we impressed upon them that we actually gave a shit. There’s a lot of politicking in this job.
Meeting 2: Branding
Later that week we had Branding discussions with all the companies. These are pretty fun, as meetings go. We ordered a keg to their office and once everyone was a couple in, started breakout sessions with the companies.
Typical structure for branding meetings is for the agency to ask a question, then go around the room and listen to all the answers. If you’re a stickler for accuracy, you might require participants to write their answers to the questions beforehand.
Typical branding questions:
—If Company X were an animal, what animal would it be?
—Circle five words that represent Company X’s brand, as it exists today: cause, invent, plan, stimulate, gamble, accomplish, respond, nurture, sex, win over [list continues…]
—What does Company X stand for?
—What reputation would you like Company X to have in three years?
And so on. The animal one is the classic branding question.
If this sounds silly, it turns out to be surprisingly useful; not least because it makes the client feel like they’ve really let loose and been heard.
Meeting 3: Web Strategy
Same as branding: four breakout meetings with the separate companies. Again, it was mostly us asking questions and listening. No keg this time.
Questions included:
—Who’s your website's target audience? Does the site need separate—the term of art is silo’d—areas for different types of clients?
—Is this site a “digital calling card,” or should it actually have some useful functionality?
—What design characteristics do you want to see? (sophisticated, playful, aspirational, refined, etc.)
—Which of your competitors have good websites? What do you like about them?
—What pages from your current site should we keep? How should we change them?
—What new pages / sections would you like?
—How much time does your company have to update the site? This one's important and overlooked way too often—every company wants a blog and a social component and Employee of the Month / Quote of the Day / Tip of the Week section that will require continual updates. 1 in 50 has the manpower to actually maintain that shit. Always flag for clients where their ideas are committing them to lots of extra work.
During initial Discovery meetings, I refrain from offering any meaningful recommendations. To wrap things up, I tell the client I've learned a ton and will get back to them soon.
Meeting 4: Tech Team
These companies had a couple in-house devs. We met with them and asked their thoughts on choice of CMS, whether they saw any roadblocks in redeveloping and deploying the websites, whether there were any other digital assets we should consider during Branding work.
This was useful for two reasons:
1) They raised a couple technical issues that later caused me to revise my estimate upwards.
2) During sales and Discovery, it’s wise to sniff out stakeholders (people on the client-side who have an interest and / or voice in the process) who don’t attend the meetings with you & your team. In big organizations, there will be a few.
I knew these developers would eventually review my development plans and pricing, so I wanted to show them I’m a decent guy who knows his shit, more or less. Again, politicking.
Meeting 5: First Draft Plans
Based on those meetings, we put together website & branding plans for all four companies. For each one I used this outline:
Website
—Key Goals: 4–5 bullet points; very general and bullshit-y. “Drive home Company X’s status as an elite HR agency”; “Provide potential clients an entry point for starting a relationship with Company X.”
—Website Sections: Same as the template list I outlined in Part 3, but more descriptive. Example: Instead of saying the homepage has a rich text block, I’d say it features “a succinct, compelling expression of Company X’s services and value props.”
Branding
—Key Brand Goals: 4-5 bullets. “Company X stands out as a company of awesome people who understand how to do X, Y, and X in a high-touch, personal style.” “Appeal to disparate clientele by walking a fine line between humor and professionalism.”
—What Company X Offers: 2-3 bullets
—What Sets Company X Apart: 2-3 bullets
—Brand Challenges: 2-3 bullets. “Must counter negative perceptions that…” “Must convey that Company X is elite and worth a premium price.”
—Branding Deliverables: What we’re gonna produce. A typical list is brand guidelines, logo, and business cards / letterhead / email signatures.
I’m a fan of sprinkling these documents with quotes lifted directly from previous meetings with the client. Put them in quotation marks to be clear about what you’re doing. It feels lazy but people love it—imagine how you’d feel if some consultant picked your words to encapsulate a key point about your company.
We had a round of face-to-face meetings to get feedback on these plans. That gave the client a chance to say “we don’t need this website feature” or “you forgot this branding deliverable”—essential to know, as we were getting close to creating a contract for the actual branding / design / dev work.
From a client relations standpoint, this was also a chance for them to ooh and ahh over how well we understood them.
Meeting 6: First Draft Plans
We revised the plans docs based on their feedback, then discussed over conference calls. Little different from the previous round.
Meeting 7: Ballpark Proposal
Created a ballpark proposal along the lines of what I described in the last post. Because of the size of the job, this monstrosity was 25+ pages. Contents:
1) Overview
2) Services for Company X1
—Website design: frontend features & content, design process
—Web copywriting
—Web backend development
—Responsive development (they had some very specific reqs here)
—Branding: Deliverables and process
3, 4, 5) Services for Company X2, X3, X4
—Same as above with minor variations as needed
6) Process
—Which meetings would be in person vs. conference calls
—Which of my staff would be involved & how. It’s common to stipulate that certain key deliverables must be created entirely by a senior designer, while others can be delegated to junior folks.
—Rough timeline, plus giant disclaimers that this could be changed later
7) Client Responsibilities
—Client had to provide current branding assets, give us internal market research docs, give us access to key personnel, pay for various 3rd-party fees
8) Payment
—Fixed price
—Milestones
—Hourly work agreement - for incidental design or dev work they might want along the way
Since it included prices, this document required a ton of prep work. A ballpark estimate isn’t binding, but you’ll look extra bush-league if you revise your numbers significantly upward between the ballpark and the actual contract (unless the client expands scope). So we approached this as we would a final draft.
My creative director and I handled estimates for each design and branding deliverable.
I had my lead developer and my best frontend and backend junior devs independently do estimates for the dev work, then had a long meeting with all of them to hash out the discrepancies in their numbers.
I then fit those hours estimates to a months-long schedule to make sure we weren’t committing to anything that exceeded the company’s bandwidth.
We presented the ballpark only to the C-level leadership of all four companies. They had some sticker shock, which we tamped down by explaining how much work is involved a project this big—clients always underestimate the difficulty of what they want.
Other than that, no real criticisms.
Meeting 8: Final Proposal
Double checked our estimates and schedule, added a signature page to the ballpark, and sent it back to client. End of discovery.
Total time elapsed: About six weeks. Could’ve been condensed 40% if the client responded faster (often the case).
4.5 Branding Phase
The good thing about a comprehensive Discovery is that, once it’s done, I can skip the kickoff discussions that would typically begin a Branding or Design project. I’ve already got all the information I need.
For this branding phase, we’d promised to prepare four things for each company
—New / ‘refreshed’ logo
—Letterhead
—Business cards
—Email signature
Normally, a rebranding project entails preparing a “brand guidelines” document which would’ve added many hours & thousands of dollars to the price. Guidelines are useful for clients because they explain all the decisions behind your branding work and give examples of how to create more branded materials. Working off the guidelines, low-cost designers can create easily create brand-consistent work.
This client decided to pass on that because the above deliverables, plus the web redesign, pretty much covered all their needs.
If I’m doing it right, the structure and approximate schedule of Branding and Design are stipulated in the contract, and that determines the flow of both phases.
Our contract promised 3 concepts of the logo & 5 revisions of the chosen logo concept, plus 2 concepts of the other deliverables and 3 revisions. My Creative Director had to take primary responsibility for all concepts and the first revision. After that, the changes requested were much smaller, and we could pass the work to a cheaper designer.
The timetable looked like this:
Meeting 1: Logo Concepts for Company X1 & X2 – Ten days after signing the contract, we gave an in person presentation of all logo concepts for the first two brands.
Meeting 2: Logo Concepts Feedback – Two days later, we had a conference call to receive feedback.
Meeting 3: Logo Revisions #1 – Two days after that, we turned around revisions of the client’s chosen concepts. Face-to-face presentation.
Meeting 4: Logo Revisions #1 Feedback – Two more days, and we got feedback on a conference call.
Meeting 5: Logo Revisions #2 & Feedback – At this point, we were making minor adjustments and the client was comfortable giving feedback immediately during the presentation meeting.
After that, we continued turning revisions around every 2-3 days until we got through the five rounds of revisions promised in the contract.
And in fact, we may have done more than five rounds—that number is specified in the contract to protect against clients making endless, unreasonable demands. I suggest not being a stickler about contractual points until it’s absolutely necessary to stay on schedule & budget.
Once the client chose a concept, we started a similar Concepts-Feedback-Revisions pattern for the letterhead, cards, and email sigs. We presented concepts and first revisions in person, then switched to conference calls. Turnaround time on revisions was similar, 2 or 3 days.
Start to finish, it took us about 5 weeks to reach final versions of all branding deliverables for the first two companies.
Then we started over and did the same process for the other two companies.
Branding on a fixed-price project is risky because what constitutes a good logo / business card / [insert other brand asset] is subjective and the designer has limitless options.
Web design is also subjective, but options are constrained by functional requirements (must have a navbar, has to work on certain screen sizes, must be buildable within client budget) and existing brand guidelines govern most of the aesthetic choices.
To protect ourselves, we spent a lot of time (gently) explaining a couple ground rules for branding:
1) We’re the experts. Our ideas are good and will work, even the client doesn’t see it right away.
2) The client is responsible for conceptual feedback (“this feels too edgy,” “needs to be more striking”) but not specific design ideas (“add some flames in the background”). If you let the client dictate by telling you what to design, the process will never end and/or the final product will be awful.
Even so, we hit a snag when the President of one of the individual companies would not get onboard with any of our logo concepts.
That’s when the early politicking becomes handy. I was buddy-buddy with the owner of all the companies, and I was able to have a frank conversation with him about the trouble this guy was causing us. He intervened and we got back on track.
4.6 Web Design Phase
Web design follows a pattern very similar to branding: Present concepts of all page templates, let the client choose, and then proceed through the allotted number of revisions until the client’s happy.
To add another layer of organization, in the contract I’ll define which templates are “primary” and “secondary.” Primary templates are designed first, and get more concepts and revisions.
The sites we built in this project were fairly basic in structure—for simplicity’s sake, let’s say each had 8 templates: homepage, basic internal, and then a “listing page” and “detail page” for three sections.
(FYI, listing & detail templates are very common in website structures. As one example, the main page of a blog is a listing page, since it lists all the posts. The page for a single blog post is a detail page. Another example: A page that shows portraits of everyone who works at a company is a listing page. Click on a portrait and you get a bio of the individual—that’s the detail page.)
For each site, I divided the primary & secondary templates like this:
Primary
—Homepage
—Basic internal
—Listing & Detail (one pair)
Secondary
—Listing & Detail (two pairs)
Primary templates got 3 concepts and 3 revisions as wireframes, and then 2 concepts and 5 revisions as mockups. See here for an example of a wireframe (right) vs. a mockup (left). Creative Director was responsible for all concepts and first revisions.
Secondary templates got 2 concepts and 3 revisions as mockups only. Creative Director only had to provide guidance on these.
The Design phase schedule looked like this:
Meeting 1: Present concepts for homepage & basic internal wireframes
Meeting 2: Feedback
Meeting 3: Present revisions of homepage & basic internal wireframes
Meeting 4: Feedback
Meeting 5:
—Present revisions of homepage & basic internal wireframes
—Present wireframe concepts of primary listing & detail page
Meeting 6: Feedback
Meeting 7:
—Present homepage & basic internal mockup concepts
—Present primary listing & detail page wireframe revisions
Meeting 8: Feedback
Meeting 9:
—Present homepage & basic internal mockup revisions
—Present primary listing & detail page mockup concepts
Meeting 10: Feedback
Meeting 11:
—Present homepage & basic internal mockup revisions
—Present primary listing & detail page mockup revisions
—Present secondary listing & detail page mockup concepts
—Feedback
(At this point we started wrapping presentations & feedback into the same meeting)
Meeting 12:
—Present homepage & basic internal mockup revisions
—Present primary listing & detail page mockup revisions
—Present secondary listing & detail page mockup revisions
—Feedback
Meeting 13:
—Present homepage & basic internal mockup revisions [final round]
—Present primary listing & detail page mockup revisions [final round]
—Present secondary listing & detail page mockup revisions
—Feedback
Meeting 14:
—Present secondary listing & detail page mockup revisions [final round]
—Feedback
And then, send over JPGs of final designs so the client has a record what they’ve agreed to.
There were likely some emails traded after that last meeting to adjust details here and there. Potentially out-of-scope as per the contract, but I’ll always honor requests like that, provided they don’t require much time.
We went through that 14-ish meeting process with two companies at a time. After the first few rounds, I would drop out and let the Creative Director and Head of Strategy take over. Later meetings were all conducted by phone.
After we finished the first two companies, we repeated with the next two.
Web design work began immediately upon wrapping a company’s branding deliverables. Similar to branding, it took about 5 weeks per site.
4.7 Copy Phase
Basically the same as web design & branding, only with a lot fewer rounds. The Head of Strategy oversaw this process.
Once the client chose a concept for a page template’s mockup, the Head of Strategy would start writing the necessary copy—or assign it to a freelancer he regularly works with.
For copy, we set the limit at one concept, two revisions. Review meetings were wrapped into the web design meetings outlined above.
Copy isn’t a standard part of web design and I’ve only done two big budget projects where it was requested. Based on that limited sample, my observation is that clients are less particular about copy than about graphic design, and you can push through the copy phase with a lot less fuss & revising.
4.8 Web Development
My wheelhouse.
Until this point, my role was limited to setting schedule & hours budgets, handling client communications, putting together the contract, sending invoices when necessary, and remaining visible / available to the client. The feedback I gave to the Creative Director and Head of Strategy was limited and very high-level (when I gave it at all).
During the Dev phase, all final responsibility rests with me.
To start web development, I’ll review the designs and tech requirements and create two schedules:
1) One schedule lists all the tasks to be accomplished during development, in the approximate order they should be tackled. Templates to code, functionalities to enable, dev / test sites to configure.
This list is extremely granular and goes on for pages, even on simple sites. Along with each task I list the hours required.
(I say “I” do this—really, it’s usually the lead developer.)
2) The second schedule shows which tasks should be finished each day. I’ll check this schedule several times per day to make sure the devs are keeping pace.
These schedules give me specific target dates for each step of development.
To the client, I’ll provide more general timelines: test site will launch [this week], QA will require [this many weeks], expect to launch [on or around this date].
For this project, all four sites were straightforward and, best of all, very similar. That meant that after we developed the first one, we could reuse large chunks of code for the others.
My estimates were at least three weeks of full-time development for the first site, and no more than two for the rest. For each site, one dev would handle both frontend & backend.
As development progressed, I would send the client a couple updates per week. Their involvement was over at this point, but you never want complete radio silence.
The rough timetable for development looked like this:
Week1: Full time dev on Company X1
Week2: Full time dev on Company X1
Week3:
—Full time dev completes Company X1
—Full time dev starts Company X2
Week4:
—Launch test site for Company X1; QA, content insertion, & client review
—Full time dev completes Company X2
—Full time dev starts Company X3
Week5:
—QA, content insertion, & client review for Company X1 wrap
—Launch test site for Company X2; QA, content insertion, & client review
—Full time dev completes Company X3
—Full time dev starts Company X4
Week 6:
—QA, content insertion, & client review for Company X2 wrap
—Launch test site for Company X3; QA, content insertion, & client review
—Full time dev completes Company X4
Week 7:
—QA, content insertion, & client review for Company X3 wrap
—Launch test site for Company X4; QA, content insertion, & client review
Week 8:
—QA, content insertion, & client review for Company X4 wrap
Week 9, 10:
—Set up hosting, analytics, webmaster tools in prep for launches
—Lead developer does final checks on our work vs. best practices
—Otherwise, holding pattern: client wants to do some press shit and internal prep ahead of simultaneous site & rebranding launch
Week 11: Launch
—We handle the website launches. Client’s responsible for deploying branding however they want.
Note that the first few weeks of this schedule overlapped with the last weeks of web design for Companies X3 & X4.
I personally do a broad-strokes internal QA on all sites we launch. I also hire QA & content insertion “specialists” to vet very fine details.
I’ll set up a few review meetings with the client to get their feedback on the test site. In theory, unless something is truly broken, my company isn’t responsible for changing it. In practice, I’m lenient on that (partly because my estimates always account for some out-of-scope client requests in the run up to launch).
I also hold training sessions for CMS-based sites and provide my standard instructional documents for Magento / WP / Drupal.
4.9 Ongoing Maintenance
Almost without exception, clients want to change or add to their site once it’s launched. Feedback from business partners or family members or employees that weren’t involved the redesigns will give them brilliant new ideas, or make them rethink stuff they thought was brilliant before.
In anticipation of these second thoughts, plus general ongoing design / dev needs, I push to set up hourly retainer agreements after the fixed price work is done. It’s easy for even small clients to burn through $1k monthly retainers.
In the project described above, the client signed a $3k retainer, which thus far they’ve used mostly for design work on posters, infographics, holiday party invitations, and other non-web stuff.
Retainers are great. My shop, even with it’s limited time in business, already gets a substantial revenue each month from its retainers. And on average retainer work is easier than initial development, and can be assigned to junior (i.e., low cost) devs and designers.
4.10 Finish
So that’s a really large project, from start to finish.
It’s important to go into any project with a clear understanding of what each phase will entail, because:
1) It helps you check that you’ve got the time and staffing for all your projects
2) When formulating estimates, you need to know how many meetings / concepts / revisions / miscellaneous tasks are required between kickoff and launch. If you don’t foresee the full complexity of every phase and each deliverable, the tendency is to underestimate and whittle away your margins.
3) Clients love to talk procedure. They’ll ask: What’s the process for Design? What’s involved in development? How long between launching the test site and launching the live site? Be ready to answer these questions like a seasoned vet, and you’ll impress + reassure them.
4) Having a plan lets you set the agenda. You should always be scheduling the next meeting and defining the next deliverable/goal. Clients hire you because they’re unfamiliar with this shit. They’re in a vulnerable, uncertain situation and they want to be led! (One among many similarities between managing clients and gaming girls.)
Of course, process varies and has to be tailored to each project. For my first few jobs, the process described here would’ve been sandblasting a cracker.
In the next—and much shorter—post, I’m going to outline a scaled-back version of the above that I use on small projects. After that I’ll do a few posts on staffing, pricing, & mechanics of day to day work.
As case studies, I’m gonna use two of my recently-launched projects: one a major rebrand / web redesign / dev project for several sister companies, and the other a much smaller design & dev project that resulted in an ‘internal prototype’ for a publishing group.
This first post tackles the bigger project and illustrates all the ins & outs / phases / considerations a project might have. I’ll walk through a very comprehensive framework for organizing a high-complexity, high-budget job.
The next post will describe the smaller project, and distill the comprehensive framework to the key pieces needed to do quality work within a limited budget and timeframe.
After these posts, I’ll give more detail on:
—hiring developers and designers
—PM best practices
—pricing
—some of my key organizational techniques.
These case studies should provide some helpful context for the points I’ll make in those later posts.
Alright. Down to business.
PART 4: The Complete Project Lifecycle
4.1 From Generalist to Specialist & Back
If you follow my career trajectory and start from the bottom, you’ll begin by handling every aspect of a project: design, development, content insertion, launch, and sometimes ongoing maintenance. You might throw in some copywriting or set up the client with a professional photographer.
You need those first clients' business, so you give them whatever services they ask.
After awhile, you’ll probably find it more profitable to specialize in one or two phases of a project.
For example, I soon began signing clients, working with them to create a written design spec, hiring a design agency to take 95% control of the project during wireframing and design, and then taking the wheel again for development.
Development was what I knew and where I earned my money, so specialized in it.
During that time, I was getting more complex projects, but I didn’t have any connections to hire designers directly. Even if I could’ve hired them, I didn’t have the experience to supervise their work. Outsourcing to a reputable agency was my best option.
The drawbacks to that were A) my company wasn’t making money off design, B) because of the agency’s markup, the clients and I were paying $150 / hour for $50 / hour junior design work, and C) I lost control during design.
Which sucked. So I reached out to design folks, tried some out, found a freelance Creative Director, and assembled a team of freelance designers directly responsible to him & me. Increasingly I cut out the design agency on projects I personally pursued / landed.
As luck had it, my Creative Director had a reputation for being an excellent web strategist (“design is strategy” is a common refrain in this industry). He also had a working relationship with a well-regarded guy who freelanced on projects as a “Head of Strategy.”
With their help, I started landing a whole new kind of client: Instead of clients who came to me with detailed specs and long lists of demands, they would pay us to A) tell them what to do and B) execute our recommendations.
The transition was akin to going from a carpenter, who works off someone else’s blueprint, to an architect who has the vision and authority to chart the creative course. Excuse the shitty analogy.
With that shift, my company was once again overseeing the entire course of projects: initial research & strategy (aka “Discovery”), design spec, design, development & launch.
4.2 Project Types
I explain all this to show that what I uniformly label as “projects” actually come in many different types. On two simultaneous engagements, my company might handle completely different kinds of work.
On some projects, we only do dev. Others, we do only design.
Some, we've done wireframes (initial outlines of designs), let another firm do final designs, and then handled dev.
Once, I ran Discovery and formulated a design & tech spec, which the client turned into an RFP and opened for bidding. I then lost the bid process and had nothing to do with rest of the project.
When I say “project lifecycle,” I’m referring to the full extent of my own shop’s involvement in the project.
As you can see, that’s sometimes only one phase, and sometimes it’s every last aspect.
4.3 The Client
The job I’m outlining here was the full-service sort. A family of four outsourced HR firms (I call them Companies X1, X2, X3, and X4 below) had retained a major NYC agency to conduct a Discovery round and propose a plan for complete rebrand and web redesign.
Assuming the plan was half decent, the agency would’ve transitioned directly into the $150k rebrand/redesign/development project.
In fact, the big agency treated the HR firms like small potatoes and gave them a boilerplate plan. HR firms balk; big agency gets fired; LouieG & Friends step in to the save the day.
The HR firms wanted my company to start from scratch. We’d run a new Discovery phase, create our own plans for branding and web redesign, and then get all the branding / design / dev work, provided our plans didn’t blow.
(No need to worry about that big agency. They picked up where they left off: Handling branding and websites for companies with multibillion dollar market caps.)
4.4 Discovery Phase
The purpose of a Discovery Phase is to A) identify specific project goals, B) create a list of content-to-create / features-to-develop that satisfies those goals, and then C) bend the content & features until they fit within the client’s budget and (when applicable) technical capacities.
The deliverable is usually some combination of:
—Website plan: A written doc including content overview, page templates, feature set & functionality, and information architecture map
—Design overview: Describing the style of web design and likely design elements; look for words like “playful,” “modern / clean,” “spacious,” “savvy,” “humorous,” “aspirational."
—Branding plan (when branding is in scope): Similar language and intent to the design overview. Also includes a list of branding assets to make: logo, brand guidelines, letterhead, business cards, email signature.
—Tech specs: CMS / hardcoded? Stripe / PayPal / Authorize.Net? Responsive web design? Retina ready graphics? If there are ambiguous functions or workflows, this is an opportunity to spell out how they’ll work, step-by-step.
Discovery isn’t a necessary part of every project. If the client comes to me knowing what they want in every detail, I skip it.
If they’re expecting me to generate ideas, I'll insist on a Discovery phase. Finding/refining ideas and getting them approved by the client is a time-intensive process that absolutely should be built into a project’s pricing and timeline.
With regard to the specific project I’m discussing—I won’t say much about contracts here, but as a note on process, we began with a $6k deal for 60 hours of discovery work. $4k on signature, $1k on delivery of rough draft plans and final draft plans.
No guarantee we’d get a dime beyond that, although we & the client were hoping the relationship worked out.
So this particular Discovery phase went as follows:
Meeting 1: Kickoff
Actually, five kickoff meetings: one with the stakeholders from all four sister companies, then breakout sessions with the individual companies.
A well-run kickoff meeting includes a lot of ‘handwaving’—business jargon for shit you do simply for appearances, to reassure the client you’re doing job. We explained the Discovery deliverables, outlined our process & the meetings we’d conduct, and spoke about our belief in the importance of doing good work and really listening to clients.
The client, for their part, bitched about their poor treatment from the previous agency, complained about their self-evidently terrible websites, and told us how they believe in doing good work and really listening to clients.
Signal-to-noise was 1:100. Kickoff meetings are like that.
The important thing was that, somewhere amidst all the platitudes, we impressed upon them that we actually gave a shit. There’s a lot of politicking in this job.
Meeting 2: Branding
Later that week we had Branding discussions with all the companies. These are pretty fun, as meetings go. We ordered a keg to their office and once everyone was a couple in, started breakout sessions with the companies.
Typical structure for branding meetings is for the agency to ask a question, then go around the room and listen to all the answers. If you’re a stickler for accuracy, you might require participants to write their answers to the questions beforehand.
Typical branding questions:
—If Company X were an animal, what animal would it be?
—Circle five words that represent Company X’s brand, as it exists today: cause, invent, plan, stimulate, gamble, accomplish, respond, nurture, sex, win over [list continues…]
—What does Company X stand for?
—What reputation would you like Company X to have in three years?
And so on. The animal one is the classic branding question.
If this sounds silly, it turns out to be surprisingly useful; not least because it makes the client feel like they’ve really let loose and been heard.
Meeting 3: Web Strategy
Same as branding: four breakout meetings with the separate companies. Again, it was mostly us asking questions and listening. No keg this time.
Questions included:
—Who’s your website's target audience? Does the site need separate—the term of art is silo’d—areas for different types of clients?
—Is this site a “digital calling card,” or should it actually have some useful functionality?
—What design characteristics do you want to see? (sophisticated, playful, aspirational, refined, etc.)
—Which of your competitors have good websites? What do you like about them?
—What pages from your current site should we keep? How should we change them?
—What new pages / sections would you like?
—How much time does your company have to update the site? This one's important and overlooked way too often—every company wants a blog and a social component and Employee of the Month / Quote of the Day / Tip of the Week section that will require continual updates. 1 in 50 has the manpower to actually maintain that shit. Always flag for clients where their ideas are committing them to lots of extra work.
During initial Discovery meetings, I refrain from offering any meaningful recommendations. To wrap things up, I tell the client I've learned a ton and will get back to them soon.
Meeting 4: Tech Team
These companies had a couple in-house devs. We met with them and asked their thoughts on choice of CMS, whether they saw any roadblocks in redeveloping and deploying the websites, whether there were any other digital assets we should consider during Branding work.
This was useful for two reasons:
1) They raised a couple technical issues that later caused me to revise my estimate upwards.
2) During sales and Discovery, it’s wise to sniff out stakeholders (people on the client-side who have an interest and / or voice in the process) who don’t attend the meetings with you & your team. In big organizations, there will be a few.
I knew these developers would eventually review my development plans and pricing, so I wanted to show them I’m a decent guy who knows his shit, more or less. Again, politicking.
Meeting 5: First Draft Plans
Based on those meetings, we put together website & branding plans for all four companies. For each one I used this outline:
Website
—Key Goals: 4–5 bullet points; very general and bullshit-y. “Drive home Company X’s status as an elite HR agency”; “Provide potential clients an entry point for starting a relationship with Company X.”
—Website Sections: Same as the template list I outlined in Part 3, but more descriptive. Example: Instead of saying the homepage has a rich text block, I’d say it features “a succinct, compelling expression of Company X’s services and value props.”
Branding
—Key Brand Goals: 4-5 bullets. “Company X stands out as a company of awesome people who understand how to do X, Y, and X in a high-touch, personal style.” “Appeal to disparate clientele by walking a fine line between humor and professionalism.”
—What Company X Offers: 2-3 bullets
—What Sets Company X Apart: 2-3 bullets
—Brand Challenges: 2-3 bullets. “Must counter negative perceptions that…” “Must convey that Company X is elite and worth a premium price.”
—Branding Deliverables: What we’re gonna produce. A typical list is brand guidelines, logo, and business cards / letterhead / email signatures.
I’m a fan of sprinkling these documents with quotes lifted directly from previous meetings with the client. Put them in quotation marks to be clear about what you’re doing. It feels lazy but people love it—imagine how you’d feel if some consultant picked your words to encapsulate a key point about your company.
We had a round of face-to-face meetings to get feedback on these plans. That gave the client a chance to say “we don’t need this website feature” or “you forgot this branding deliverable”—essential to know, as we were getting close to creating a contract for the actual branding / design / dev work.
From a client relations standpoint, this was also a chance for them to ooh and ahh over how well we understood them.
Meeting 6: First Draft Plans
We revised the plans docs based on their feedback, then discussed over conference calls. Little different from the previous round.
Meeting 7: Ballpark Proposal
Created a ballpark proposal along the lines of what I described in the last post. Because of the size of the job, this monstrosity was 25+ pages. Contents:
1) Overview
2) Services for Company X1
—Website design: frontend features & content, design process
—Web copywriting
—Web backend development
—Responsive development (they had some very specific reqs here)
—Branding: Deliverables and process
3, 4, 5) Services for Company X2, X3, X4
—Same as above with minor variations as needed
6) Process
—Which meetings would be in person vs. conference calls
—Which of my staff would be involved & how. It’s common to stipulate that certain key deliverables must be created entirely by a senior designer, while others can be delegated to junior folks.
—Rough timeline, plus giant disclaimers that this could be changed later
7) Client Responsibilities
—Client had to provide current branding assets, give us internal market research docs, give us access to key personnel, pay for various 3rd-party fees
8) Payment
—Fixed price
—Milestones
—Hourly work agreement - for incidental design or dev work they might want along the way
Since it included prices, this document required a ton of prep work. A ballpark estimate isn’t binding, but you’ll look extra bush-league if you revise your numbers significantly upward between the ballpark and the actual contract (unless the client expands scope). So we approached this as we would a final draft.
My creative director and I handled estimates for each design and branding deliverable.
I had my lead developer and my best frontend and backend junior devs independently do estimates for the dev work, then had a long meeting with all of them to hash out the discrepancies in their numbers.
I then fit those hours estimates to a months-long schedule to make sure we weren’t committing to anything that exceeded the company’s bandwidth.
We presented the ballpark only to the C-level leadership of all four companies. They had some sticker shock, which we tamped down by explaining how much work is involved a project this big—clients always underestimate the difficulty of what they want.
Other than that, no real criticisms.
Meeting 8: Final Proposal
Double checked our estimates and schedule, added a signature page to the ballpark, and sent it back to client. End of discovery.
Total time elapsed: About six weeks. Could’ve been condensed 40% if the client responded faster (often the case).
4.5 Branding Phase
The good thing about a comprehensive Discovery is that, once it’s done, I can skip the kickoff discussions that would typically begin a Branding or Design project. I’ve already got all the information I need.
For this branding phase, we’d promised to prepare four things for each company
—New / ‘refreshed’ logo
—Letterhead
—Business cards
—Email signature
Normally, a rebranding project entails preparing a “brand guidelines” document which would’ve added many hours & thousands of dollars to the price. Guidelines are useful for clients because they explain all the decisions behind your branding work and give examples of how to create more branded materials. Working off the guidelines, low-cost designers can create easily create brand-consistent work.
This client decided to pass on that because the above deliverables, plus the web redesign, pretty much covered all their needs.
If I’m doing it right, the structure and approximate schedule of Branding and Design are stipulated in the contract, and that determines the flow of both phases.
Our contract promised 3 concepts of the logo & 5 revisions of the chosen logo concept, plus 2 concepts of the other deliverables and 3 revisions. My Creative Director had to take primary responsibility for all concepts and the first revision. After that, the changes requested were much smaller, and we could pass the work to a cheaper designer.
The timetable looked like this:
Meeting 1: Logo Concepts for Company X1 & X2 – Ten days after signing the contract, we gave an in person presentation of all logo concepts for the first two brands.
Meeting 2: Logo Concepts Feedback – Two days later, we had a conference call to receive feedback.
Meeting 3: Logo Revisions #1 – Two days after that, we turned around revisions of the client’s chosen concepts. Face-to-face presentation.
Meeting 4: Logo Revisions #1 Feedback – Two more days, and we got feedback on a conference call.
Meeting 5: Logo Revisions #2 & Feedback – At this point, we were making minor adjustments and the client was comfortable giving feedback immediately during the presentation meeting.
After that, we continued turning revisions around every 2-3 days until we got through the five rounds of revisions promised in the contract.
And in fact, we may have done more than five rounds—that number is specified in the contract to protect against clients making endless, unreasonable demands. I suggest not being a stickler about contractual points until it’s absolutely necessary to stay on schedule & budget.
Once the client chose a concept, we started a similar Concepts-Feedback-Revisions pattern for the letterhead, cards, and email sigs. We presented concepts and first revisions in person, then switched to conference calls. Turnaround time on revisions was similar, 2 or 3 days.
Start to finish, it took us about 5 weeks to reach final versions of all branding deliverables for the first two companies.
Then we started over and did the same process for the other two companies.
Branding on a fixed-price project is risky because what constitutes a good logo / business card / [insert other brand asset] is subjective and the designer has limitless options.
Web design is also subjective, but options are constrained by functional requirements (must have a navbar, has to work on certain screen sizes, must be buildable within client budget) and existing brand guidelines govern most of the aesthetic choices.
To protect ourselves, we spent a lot of time (gently) explaining a couple ground rules for branding:
1) We’re the experts. Our ideas are good and will work, even the client doesn’t see it right away.
2) The client is responsible for conceptual feedback (“this feels too edgy,” “needs to be more striking”) but not specific design ideas (“add some flames in the background”). If you let the client dictate by telling you what to design, the process will never end and/or the final product will be awful.
Even so, we hit a snag when the President of one of the individual companies would not get onboard with any of our logo concepts.
That’s when the early politicking becomes handy. I was buddy-buddy with the owner of all the companies, and I was able to have a frank conversation with him about the trouble this guy was causing us. He intervened and we got back on track.
4.6 Web Design Phase
Web design follows a pattern very similar to branding: Present concepts of all page templates, let the client choose, and then proceed through the allotted number of revisions until the client’s happy.
To add another layer of organization, in the contract I’ll define which templates are “primary” and “secondary.” Primary templates are designed first, and get more concepts and revisions.
The sites we built in this project were fairly basic in structure—for simplicity’s sake, let’s say each had 8 templates: homepage, basic internal, and then a “listing page” and “detail page” for three sections.
(FYI, listing & detail templates are very common in website structures. As one example, the main page of a blog is a listing page, since it lists all the posts. The page for a single blog post is a detail page. Another example: A page that shows portraits of everyone who works at a company is a listing page. Click on a portrait and you get a bio of the individual—that’s the detail page.)
For each site, I divided the primary & secondary templates like this:
Primary
—Homepage
—Basic internal
—Listing & Detail (one pair)
Secondary
—Listing & Detail (two pairs)
Primary templates got 3 concepts and 3 revisions as wireframes, and then 2 concepts and 5 revisions as mockups. See here for an example of a wireframe (right) vs. a mockup (left). Creative Director was responsible for all concepts and first revisions.
Secondary templates got 2 concepts and 3 revisions as mockups only. Creative Director only had to provide guidance on these.
The Design phase schedule looked like this:
Meeting 1: Present concepts for homepage & basic internal wireframes
Meeting 2: Feedback
Meeting 3: Present revisions of homepage & basic internal wireframes
Meeting 4: Feedback
Meeting 5:
—Present revisions of homepage & basic internal wireframes
—Present wireframe concepts of primary listing & detail page
Meeting 6: Feedback
Meeting 7:
—Present homepage & basic internal mockup concepts
—Present primary listing & detail page wireframe revisions
Meeting 8: Feedback
Meeting 9:
—Present homepage & basic internal mockup revisions
—Present primary listing & detail page mockup concepts
Meeting 10: Feedback
Meeting 11:
—Present homepage & basic internal mockup revisions
—Present primary listing & detail page mockup revisions
—Present secondary listing & detail page mockup concepts
—Feedback
(At this point we started wrapping presentations & feedback into the same meeting)
Meeting 12:
—Present homepage & basic internal mockup revisions
—Present primary listing & detail page mockup revisions
—Present secondary listing & detail page mockup revisions
—Feedback
Meeting 13:
—Present homepage & basic internal mockup revisions [final round]
—Present primary listing & detail page mockup revisions [final round]
—Present secondary listing & detail page mockup revisions
—Feedback
Meeting 14:
—Present secondary listing & detail page mockup revisions [final round]
—Feedback
And then, send over JPGs of final designs so the client has a record what they’ve agreed to.
There were likely some emails traded after that last meeting to adjust details here and there. Potentially out-of-scope as per the contract, but I’ll always honor requests like that, provided they don’t require much time.
We went through that 14-ish meeting process with two companies at a time. After the first few rounds, I would drop out and let the Creative Director and Head of Strategy take over. Later meetings were all conducted by phone.
After we finished the first two companies, we repeated with the next two.
Web design work began immediately upon wrapping a company’s branding deliverables. Similar to branding, it took about 5 weeks per site.
4.7 Copy Phase
Basically the same as web design & branding, only with a lot fewer rounds. The Head of Strategy oversaw this process.
Once the client chose a concept for a page template’s mockup, the Head of Strategy would start writing the necessary copy—or assign it to a freelancer he regularly works with.
For copy, we set the limit at one concept, two revisions. Review meetings were wrapped into the web design meetings outlined above.
Copy isn’t a standard part of web design and I’ve only done two big budget projects where it was requested. Based on that limited sample, my observation is that clients are less particular about copy than about graphic design, and you can push through the copy phase with a lot less fuss & revising.
4.8 Web Development
My wheelhouse.
Until this point, my role was limited to setting schedule & hours budgets, handling client communications, putting together the contract, sending invoices when necessary, and remaining visible / available to the client. The feedback I gave to the Creative Director and Head of Strategy was limited and very high-level (when I gave it at all).
During the Dev phase, all final responsibility rests with me.
To start web development, I’ll review the designs and tech requirements and create two schedules:
1) One schedule lists all the tasks to be accomplished during development, in the approximate order they should be tackled. Templates to code, functionalities to enable, dev / test sites to configure.
This list is extremely granular and goes on for pages, even on simple sites. Along with each task I list the hours required.
(I say “I” do this—really, it’s usually the lead developer.)
2) The second schedule shows which tasks should be finished each day. I’ll check this schedule several times per day to make sure the devs are keeping pace.
These schedules give me specific target dates for each step of development.
To the client, I’ll provide more general timelines: test site will launch [this week], QA will require [this many weeks], expect to launch [on or around this date].
For this project, all four sites were straightforward and, best of all, very similar. That meant that after we developed the first one, we could reuse large chunks of code for the others.
My estimates were at least three weeks of full-time development for the first site, and no more than two for the rest. For each site, one dev would handle both frontend & backend.
As development progressed, I would send the client a couple updates per week. Their involvement was over at this point, but you never want complete radio silence.
The rough timetable for development looked like this:
Week1: Full time dev on Company X1
Week2: Full time dev on Company X1
Week3:
—Full time dev completes Company X1
—Full time dev starts Company X2
Week4:
—Launch test site for Company X1; QA, content insertion, & client review
—Full time dev completes Company X2
—Full time dev starts Company X3
Week5:
—QA, content insertion, & client review for Company X1 wrap
—Launch test site for Company X2; QA, content insertion, & client review
—Full time dev completes Company X3
—Full time dev starts Company X4
Week 6:
—QA, content insertion, & client review for Company X2 wrap
—Launch test site for Company X3; QA, content insertion, & client review
—Full time dev completes Company X4
Week 7:
—QA, content insertion, & client review for Company X3 wrap
—Launch test site for Company X4; QA, content insertion, & client review
Week 8:
—QA, content insertion, & client review for Company X4 wrap
Week 9, 10:
—Set up hosting, analytics, webmaster tools in prep for launches
—Lead developer does final checks on our work vs. best practices
—Otherwise, holding pattern: client wants to do some press shit and internal prep ahead of simultaneous site & rebranding launch
Week 11: Launch
—We handle the website launches. Client’s responsible for deploying branding however they want.
Note that the first few weeks of this schedule overlapped with the last weeks of web design for Companies X3 & X4.
I personally do a broad-strokes internal QA on all sites we launch. I also hire QA & content insertion “specialists” to vet very fine details.
I’ll set up a few review meetings with the client to get their feedback on the test site. In theory, unless something is truly broken, my company isn’t responsible for changing it. In practice, I’m lenient on that (partly because my estimates always account for some out-of-scope client requests in the run up to launch).
I also hold training sessions for CMS-based sites and provide my standard instructional documents for Magento / WP / Drupal.
4.9 Ongoing Maintenance
Almost without exception, clients want to change or add to their site once it’s launched. Feedback from business partners or family members or employees that weren’t involved the redesigns will give them brilliant new ideas, or make them rethink stuff they thought was brilliant before.
In anticipation of these second thoughts, plus general ongoing design / dev needs, I push to set up hourly retainer agreements after the fixed price work is done. It’s easy for even small clients to burn through $1k monthly retainers.
In the project described above, the client signed a $3k retainer, which thus far they’ve used mostly for design work on posters, infographics, holiday party invitations, and other non-web stuff.
Retainers are great. My shop, even with it’s limited time in business, already gets a substantial revenue each month from its retainers. And on average retainer work is easier than initial development, and can be assigned to junior (i.e., low cost) devs and designers.
4.10 Finish
So that’s a really large project, from start to finish.
It’s important to go into any project with a clear understanding of what each phase will entail, because:
1) It helps you check that you’ve got the time and staffing for all your projects
2) When formulating estimates, you need to know how many meetings / concepts / revisions / miscellaneous tasks are required between kickoff and launch. If you don’t foresee the full complexity of every phase and each deliverable, the tendency is to underestimate and whittle away your margins.
3) Clients love to talk procedure. They’ll ask: What’s the process for Design? What’s involved in development? How long between launching the test site and launching the live site? Be ready to answer these questions like a seasoned vet, and you’ll impress + reassure them.
4) Having a plan lets you set the agenda. You should always be scheduling the next meeting and defining the next deliverable/goal. Clients hire you because they’re unfamiliar with this shit. They’re in a vulnerable, uncertain situation and they want to be led! (One among many similarities between managing clients and gaming girls.)
Of course, process varies and has to be tailored to each project. For my first few jobs, the process described here would’ve been sandblasting a cracker.
In the next—and much shorter—post, I’m going to outline a scaled-back version of the above that I use on small projects. After that I’ll do a few posts on staffing, pricing, & mechanics of day to day work.