Web development shops

The company shape

Web development shops split into three shapes and the shapes rarely compete for the same work. The first shape is the boutique custom shop: 6 to 25 developers plus a designer or two and a project manager, revenue between $1M and $6M, project fees running $40K to $350K for a marketing site rebuild and $200K to $1.5M for a custom application. The second shape is the platform specialist: 8 to 60 people focused on a single platform (Shopify Plus, HubSpot CMS, Contentful, Webflow, Sanity, Drupal), revenue between $2M and $18M, with retainer support after project delivery. The third shape is the mid-market digital agency doing web development as one of several disciplines: 30 to 150 people, revenue between $6M and $40M, web-dev revenue running 40% to 70% of the mix.

The founding story usually starts with two developers or a developer and a designer who leave a larger agency after four to seven years, take a client relationship with them, and open a shop. Growth after the first year depends on whether the founders can hire senior engineering talent in a market where senior engineers earn more inside product companies than most agencies pay.

Utilization economics are unforgiving. The billable target sits at 65% to 75% of engineering capacity, and the reality lands 5 to 15 points lower once you count meetings, learning, sales support, and bench time between projects. Engineering rate cards run $125 to $275 per hour for senior developers, $85 to $175 for mid-level, $55 to $110 for junior. Blended shop rates run $135 to $225 depending on positioning strength.

Revenue is lumpy because project work dominates for most shops. A shop running $4M annualized sees quarterly swings from $600K to $1.4M depending on when large builds kick off. Cash management dominates the operations conversation. Bench management dominates the people conversation. The healthiest shops have a retainer layer (support, hosting, monitoring, ongoing enhancement) that produces 25% to 40% of revenue on a recurring basis and covers fixed costs during project gaps.

The organization above 40 people starts to look different: a dedicated business development function, a head of engineering, structured practice leads per platform, a defined delivery methodology that outlives any one project manager, and a formal talent development program. Shops below 15 people run on founder heroics and shared Notion pages. Shops in the 15 to 40 range live in the awkward middle where the founder-led model is straining and the mid-market shop scaffolding has not yet arrived.

The buyer

The buyer for a web development project sits in one of three seats. First, a marketing leader (VP Marketing, CMO, Head of Growth) sponsoring a marketing site rebuild inside a company with revenue between $10M and $500M. Second, a product or engineering leader (CTO, VP Engineering, Head of Product) sponsoring a custom application or a platform build. Third, an ecommerce leader (Head of Ecommerce, Director of Digital) sponsoring a Shopify Plus or headless commerce build inside a DTC or B2B commerce brand.

The three buyers have different budgets, different timelines, and different definitions of success. The marketing buyer holds discretionary budget between $50K and $500K, wants a launch inside four to six months, measures success by lead volume and marketing team velocity post-launch. The product buyer holds capital budget between $200K and $2M, expects a nine to eighteen month engagement, measures success by feature delivery and engineering team enablement. The ecommerce buyer holds mixed budget between $80K and $700K, wants launch tied to a specific commercial event, measures success by conversion rate and revenue per session.

Every buyer has been burned before. Web development projects have a long history of overrunning timelines, blowing budgets, and shipping something that does not match the pitch. Every experienced buyer has at least one scar. The pitch that lands is specific, honest about timelines, and clear about who is actually building the work. The pitch that loses promises unrealistic delivery dates or sells senior talent that turns into junior talent after the SOW is signed.

The buyer's research is thorough. They review the shop's portfolio in detail, click through to live sites, run the sites through PageSpeed Insights and Lighthouse, check the shop's own site on both, and often ask for two or three reference conversations before signing. Sophisticated buyers check the shop's GitHub organization, the technical blog posts, and the engineers' individual LinkedIn profiles.

The buyer is influenced by three groups. First, peers who have used a specific shop on a comparable project. Peer referrals close somewhere between 50% and 70% of the time in this category, higher than most agency categories because the risk of a bad choice is more visible. Second, agencies that partner with the shop (design agencies, brand agencies, marketing agencies that need an implementation partner). Third, the shop's own public work: portfolio, case studies, technical writing, open source contributions.

The buyer expects that a web development shop should ship a fast, well-built, well-maintained site of its own. Slow LCP, broken navigation, outdated content, or missing SSL cost the deal before the first call. The shop's own site is a live proof of what the shop can deliver.

Discovery landscape

Discovery for web development shops runs across a specific set of surfaces that reward technical proof and portfolio depth. The buyer's research is granular and skeptical, so the surfaces have to carry real evidence rather than marketing polish.

Google organic matters for platform-specific and use-case-specific queries. "Shopify Plus development agency," "headless Contentful build," "custom Laravel development shop," "HubSpot CMS Hub developer." These queries carry high intent and moderate volume. Shops that rank for the specific platform and use-case combinations they serve get onto shortlists during the research window.

Platform partner directories are the single most productive category-specific surface. Shopify Plus Partner directory, HubSpot Solutions Partner directory, Contentful Solution Partner, Adobe Solution Partner, Sanity Partner, Webflow Enterprise Partner. Buyers use these directories as their first filter. Elite or top-tier partner status inside the directory is worth more than any single piece of thought leadership.

Award platforms have unusual weight in this category. Awwwards, CSS Design Awards, FWA, Site Inspire, and the Webby Awards. Buyers who value design quality use these as trust signals. Awards live on the shop's site and on the individual designers' portfolios for years and produce long-tail inbound.

GitHub, technical blogs, and open source contributions matter for the shops selling into product and engineering buyers. A shop with a maintained open source library used by the target buyer's engineering team enters the conversation with credibility that no case study can produce.

Portfolio and case study depth is the largest single surface. Case studies with process detail, technical stack, performance numbers, and (where possible) business outcome numbers carry real weight. Portfolios that stop at "here is a nice screenshot" get skipped.

LinkedIn is the personal-brand distribution surface. Named principals and technical leads with a real cadence drive inbound. Named designers with a following inside the design community drive inbound. Engineering managers with published talks and conference credibility drive inbound.

AI answer engines are a growing surface. Buyers ask Claude or Perplexity to shortlist shops for specific platforms and use cases. Shops with substantive published work structured for AI answer engines get cited.

Design and development communities matter as a slow long-tail channel. Dribbble for design shops, Frontend Masters and Egghead for frontend engineering credibility, Laracasts for Laravel shops, and the equivalent community for whichever platform the shop specializes in. Presence in these communities builds credibility that flows into inbound over years.

What does not matter much: display advertising, most cold outbound, generic gated whitepapers, sponsored placements on non-technical publications. Buyers do not engage with these formats at this consideration level.

What breaks most often

1. The shop's own site is a broken advertisement

Slow LCP, layout shift, outdated content, unmaintained portfolio, and a stale technical blog. The shop sells fast, well-built sites. The buyer arrives, runs the shop's own site through Lighthouse, gets a 42, and closes the tab. Nothing else the shop publishes matters until this problem is fixed. This is the honest tension of the category, and it is fixable within a quarter.

2. Portfolio is thin and undated

The portfolio shows six projects. Two are from 2019 and use technologies the shop no longer works in. One is a project the shop bailed on halfway through. There is no case study, no process, no outcome numbers, no technical stack disclosure. The buyer's due diligence stops here.

3. Positioning is technology-broad

The site says "we build custom web applications and marketing sites in modern technologies." Every competitor says the same thing. Meanwhile the shop actually does 70% of revenue in headless Shopify Plus builds for DTC brands over $10M. The site should say that. Broad positioning attracts unqualified inbound that wastes senior time on unqualified pitches.

4. Sales collateral does not match delivery reality

The pitch deck introduces the senior architect and the design lead. The kickoff meeting after signing introduces a mid-level developer and a junior designer. The client notices. Client satisfaction erodes from week one. This is one of the largest single drivers of client dissatisfaction in the category, and it is fixable through better sales-to-delivery handoff and honest resource commitment during pitch.

5. Scope creeps and margin collapses

The SOW says four months and $180K. The project delivers seven months later at $260K of actual work billed. The client feels the schedule slip. The shop feels the margin loss. Nobody wins. Structured change management (documented change orders, weekly scope check-ins, transparent burn-down reporting) protects both the client relationship and the shop economics.

6. Post-launch support is undefined

The site launches, the project team moves to the next engagement, and nobody at the shop owns the client relationship. A month later something breaks, the client emails a shared inbox, and nobody responds for four days. The retention opportunity disappears. Structured post-launch support (a defined support retainer offer, a named account owner, a monthly check-in) turns launched projects into recurring revenue for years.

7. Platform partner status sits under-invested

The shop is a Shopify Partner but not a Shopify Plus Partner. Or a HubSpot Solutions Partner but not Elite. Or a Contentful Solution Partner but not tier one. Every partner-status tier the shop misses cuts inbound from the platform's directory. The certifications and reference client requirements take work but are the highest single-quarter ROI investment in the category.

The Ranking Surfaces Playbook applied

The Playbook applies to web development shops with two category-specific inversions. First, the shop's own site performance is a hard credibility gate. Second, platform partner directory status is a first-order sales asset that most agencies underweight. Everything downstream depends on these two foundations.

Tier one: the surfaces that produce SOWs this quarter

CWV as sales asset. The shop's own Core Web Vitals score sits in the green across desktop and mobile. Lighthouse performance score above 95. This is the first thing a buyer checks. It cannot be delegated to later. For any other agency category CWV sits at tier three; for web development it belongs at tier one.

Platform partner directory presence. Elite or top-tier status inside the platform partner program for the shop's specialization. This includes the certifications, the reference client requirements, and the partner-portal profile management. The directory is where buyers arrive first.

Portfolio and case study depth. Twelve to twenty published case studies with process detail, technical stack disclosure, performance numbers, and business outcome numbers where the client permits. Every case study links to a live URL. Every case study has a named architect and named designer credited.

E-E-A-T for the shop and the principals. Named principal bios at 1,500 to 2,500 words with representative engagements, published technical writing, conference speaking, open source contributions, and a real point of view. Author schema on every published piece. Real About page with founding story and technical philosophy.

SEO for platform and use-case queries. Ranking for "Shopify Plus development agency," "headless Contentful build," "custom Laravel shop [region]," or whichever platform and use-case combination the shop specializes in. This is where the buyer's Google search lands.

Tier two: the surfaces that compound

AEO and GEO. Long-form pieces on platform-specific technical topics structured for AI answer engines. Direct-answer TL;DRs. FAQPage schema on subheads. Comparison tables. AI answer engines increasingly cite technical shops on specific platform decisions and architecture questions.

LinkedIn and technical community distribution. Named principals posting substantive content on LinkedIn. Named engineers with GitHub presence, technical blog posts, and community contributions inside the shop's platform.

KGO for the shop and named principals. Wikidata entries where notability supports them, Knowledge Panel work, sameAs across LinkedIn, GitHub, and speaker profiles.

Design and development award platforms. Awwwards, CSS Design Awards, FWA, Site Inspire, Webbys. Slow compounding but durable.

Tier three: worth doing, lower ROI

VxSO as a real investment. Portfolio work is visual by nature. Case study screenshots, process artifacts, before-and-after visuals, properly schematized with ImageObject.

Reputation platforms. Clutch, DesignRush, Sortlist, and Google reviews. Higher weight for younger shops without long-term brand equity, lower weight for established shops with strong partner directory presence.

VSO low. Speakable schema on FAQ as AEO free-rider.

Tier four: not a fit

ASO, GLOBO, Web3. Web development shops do not have their own apps, do not compete on international search at this size, and Web3 shops are a separate category with different economics.

LSO minor. Some shops serve a regional market and benefit from Google Business Profile hygiene. Most sell nationally and can skip.

AAO now meaningful. Deploy llms.txt v2 as a first-mover play. Watch AI answer engine citation share for named principals as a leading indicator.

The honest note. The shop's own site performance is the trust gate. Nothing else recovers a broken first impression once the buyer has run Lighthouse.

First 30 / 60 / 90 days

Days 1 to 30: fix the trust gate and lock positioning

Site performance sprint against the shop's own site. Lighthouse audit, CWV remediation plan, image optimization, third-party script pruning, hosting review. Target: green CWV across desktop and mobile inside four weeks. This is the trust gate and it cannot be deferred.

Interviews with founding principals on real practice depth. Which platform or use-case combinations produce 70% of revenue. Which industries define the ideal client. Which named engineers or designers drive the largest share of inbound. Positioning locks around the answers.

Portfolio and case study audit. Which existing case studies are current, which need to be re-approached for real numbers, which should be pulled. Platform partner status audit: what tier the shop currently holds, what the next tier requires, what certifications are missing.

Deliverable at day 30: green CWV on the shop's own site, a positioning document, a portfolio rebuild scope, a partner directory promotion plan (which tier the shop is chasing next, what work it takes), a content plan for the next quarter, and a lead-source tracking system.

Days 31 to 60: publish, promote, and prove

The first three case studies rebuild ship, each with process detail, technical stack disclosure, performance numbers, and business outcome numbers where the client permits. The first three long-form technical pieces publish, authored by named principals or engineers, structured for AEO with direct-answer TL;DRs and FAQPage schema.

Platform partner promotion work: certifications, reference client requests, partner-portal profile updates. Award submissions to Awwwards, CSS Design Awards, or the platform's own award programs where relevant.

Named principal LinkedIn cadence begins in earnest, with ghostwriting support for principals who cannot sustain the cadence themselves. Named principal bios rewritten at 1,500 to 2,500 words.

Deliverable at day 60: three rebuilt case studies, three published long-form technical pieces, refreshed principal bios, live LinkedIn cadence, active partner-directory work, and submitted award entries.

Days 61 to 90: measure and iterate

Lead-source tracking review: which inbound came from platform partner directory, Google organic, LinkedIn, awards, referrals, and AI answer engine citations. Content engagement analysis. Partner directory inbound signal (impressions inside the directory, click-throughs to the shop's profile, inbound RFQs sourced through the platform).

Post-launch support offering rebuilt: defined support retainer, named account owner, monthly check-in cadence. Existing launched projects get re-approached with the new support offering. Retention benefit compounds over months.

Deliverable at day 90: measurable ranking movement on platform and use-case queries, measurable inbound signal from published content and partner directory, active post-launch support conversations with prior clients, and a clear roadmap for months four through twelve.