Top 10 App Development Companies in Los Angeles 2026:#1 Frenchy Digital

"I searched for the best app developers in Los Angeles. The first list I opened ranked a company in New Jersey at number one. The second one ranked Accenture and IBM. Who here actually builds apps?"

That's a paraphrase of a question founders ask me, and it's a fair one. So I went and checked. One of the most visible 2026 Los Angeles lists is published by Chop Dawg, which ranks itself first and describes itself as fully distributed, serving LA remotely. Another is published by Bitcot, which also ranks itself first and fills most of the rest with global consultancies like Accenture, Infosys and SAP.

Neither list is lying, exactly. They're just answering a different question from the one you asked.

So here is my claim, with the date on it. On September 29, 2026 I scored ten app development companies that show a Los Angeles area presence on their own websites, on six things a buyer can check from a public page in about ten minutes each. TechAhead came first with 11 of 12. The average was 6.7.

To be clear about the conflict up front: I run Frenchy Digital, which builds apps in Los Angeles. It is not in the scored table. It gets its own disclosed section near the end, where I score it on the same rubric (11 of 12, which would tie TechAhead and place first on the alphabetical tiebreak) and explain why it's still my pick for a specific kind of buyer, and who should pick someone else.

If you're earlier than this and still deciding what questions to ask any agency, start with the ten questions I'd use to vet a Los Angeles app agency. This article assumes you want a shortlist of real local firms that you can defend to whoever signs the check.

Most Los Angeles app developer rankings fail for one of three reasons: the author is a contestant, the firms aren't actually in Los Angeles, or the scores rest on numbers nobody outside the company can check.

The first problem is obvious once you see it. Two of the lists that surfaced first for me put their own author at number one. That's common across the industry, and I'm not above it either, which is why my own company sits outside the table here.

The second problem is sneakier. "Los Angeles" on a ranking site often means "has a client in Los Angeles" or "bought a local landing page". I found one firm that three separate lists call an LA studio whose own website listed offices in Pleasanton, Pittsburgh, Islamabad and Shenzhen, and nothing here.

The third problem is the one that makes most rankings useless for a decision. A list that credits an agency with "500+ apps launched" or "25 million users" is repeating the agency's homepage. There is no neutral registry of app agency outcomes. Nobody audits those counters.

The wrong model is that a ranking measures who builds the best apps. The real model is that a public ranking can only measure what companies disclose. Those are different things, and pretending otherwise is how you end up shortlisting on copywriting.

Why does that matter so much for an app specifically? Because the two things that decide whether an app project goes well for you, the price structure and who owns the code at the end, are exactly the two things agencies least like to publish. You'll see that in the scoreboard in a minute.

Before scoring anyone, each candidate had to pass one test: a Los Angeles area presence shown on the company's own website, plus no sign of an acquisition, merger or shutdown from 2024 to 2026.

"Los Angeles area" here means what a buyer would reasonably drive to: the city itself, Santa Monica, Venice, Culver City, Beverly Hills, Manhattan Beach, Agoura Hills and the rest of the county. A directory profile didn't count. A ranking site saying so didn't count. The company had to say it.

I started from about 16 names that recur across LA lists and search results, then cut. Here's who didn't make it and why.

• CitrusBits: Called a long-running Los Angeles studio on several lists. Its own site listed Pleasanton, Pittsburgh, Islamabad and Shenzhen on September 29, 2026, and no LA office. Out.
• Chop Dawg: Describes itself as fully distributed and serving LA remotely, and authored an LA ranking with itself first. Out on the office test, and I don't score list authors against their own lists anyway.
• Bitcot: Authored an LA ranking with itself first. I didn't score it, for the same reason.
• Your App Hero: A directory lists a Pasadena address, but its own homepage showed a 917 New York number and no address. Out.
• Global consultancies: Accenture, IBM, Infosys and similar firms have LA staff, but they aren't app studios a founder can hire directly for an MVP. Out of scope.

On acquisitions: I searched each of the ten finalists for acquisition, merger or shutdown news from 2024 to 2026 and found none. That's a clean result, and I want to be careful with it. Absence from search results isn't proof of independence. If ownership matters to you (it should, because a new parent can change who staffs your project and which contract you sign), ask directly.

One firm in the final ten, Distillery, passes on a technicality I want to flag. Its contact page lists a Manhattan Beach headquarters address with a PMB number, which is a private mailbox. That still shows an LA County presence, so it passed the entry test. It scores 1, not 2, on the office attribute below.

Six attributes, 0 to 2 points each, twelve maximum, all checked on September 29, 2026 from pages anyone can open.

• Verifiable LA office: 2 for a street address in the LA area on the company's own site; 1 for Los Angeles named, an LA phone number or a mailbox address without an office; 0 for none.
• Pricing transparency: 2 for figures on the company's own site; 1 for a third-party rate band only; 0 for neither.
• Named app work: 2 for three or more named projects the site presents as apps; 1 for one or two, or named clients where app work is unclear; 0 for none.
• Public IP terms: 2 for an explicit source code and IP ownership statement; 1 for IP protection language; 0 for nothing stated.
• Native iOS and Android evidence: 2 if both are named as capabilities on the company's own site; 1 if mobile is only implied or shown in case study tags; 0 if not stated on the pages checked.
• Years operating: 2 for a founding year or ten-plus years stated on the company's own site; 1 for a founding year from a third-party profile only; 0 for none found.

Ties break on named app work, then years operating, then alphabetically. That's the whole rubric.

What I excluded, and why.Download counts, user totals, funding raised by clients, app store features and outcome percentages, because every one is the agency's own claim. Team size, because headcount says nothing about who works on your project. Awards, because most are paid or self-nominated and none measures delivery.

Clutch, and why it's not scored. I recorded whether I saw a Clutch listing, but I didn't score ratings or review counts. Clutch's Los Angeles app developer directoryblocks automated readers, so counts reach me only through search snippets and the companies' own badges, and they don't agree with each other. A number I can't see on the source page isn't a number I'll rank on.

How to re-check it.Open each company's contact and about pages, look for a street address and a founding year, search the site for "pricing", "source code" and "intellectual property", and look at the portfolio for named apps. It takes about ten minutes per firm. If a company has published something since September 29, its score should change, and I'd want to know.

TechAhead leads with 11 of 12. Below it the spread is moderate, and three firms tie on 6 in the middle.

Do the arithmetic and a pattern shows up. The ten scores add to 67, so the average is 6.7 of 12, about 56%. The leader, at 11, discloses roughly 1.6x what the average firm does. Three firms sit on exactly 6, which means the middle of this table is decided by tiebreaks, not by any meaningful gap.

Now read the columns instead of the rows. Pricing: seven of ten score zero. IP terms: nine of ten score zero. Every firm except one scores at least 1 on office, named work and years. So the market is fairly open about where it is and who it has worked for, and quiet about what it costs and what you own.

That's the finding I'd take away even if you ignore the ranks. Whoever you shortlist, you'll almost certainly have to ask for price and IP terms in writing.

Every score above traces to a cell below. Where a cell says "company's claim", I saw it on the company's site, not in an independent source.

A few cells deserve a note. Goji Labs' Wilshire Boulevard address comes from the structured data on its own homepage rather than a visible contact page (its /contact URL returned a 404 when I checked). It's still the company's own statement, so it scores 2.

ISBX's founding year of 2009 appears only on third-party company profiles; its homepage has a "years in business" counter that rendered as zero to my reader because it animates in with JavaScript. So it scores 1 on years. If you open the page in a browser, you'll likely see the number.

And TechAhead's native mobile score is a 1, not a 2, for a narrow reason: its homepage names mobile and product engineering, but on the pages I checked it didn't split out iOS and Android as named capabilities. It's widely known as an app firm. The rubric scores what the page shows.

Each profile says who the firm suits, what I could verify, and what I couldn't.None of this is a quality judgment. I haven't worked with any of them.

Top ranked

1

TechAhead

published prices and a code ownership promise

Score: 11 of 12. Suits: funded startups and mid-market companies that want an established firm with a written price range and a stated IP position before the first call.

Verifiable: its contact page gives 28720 Roadside Drive in Agoura Hills. Its about page says it has handled projects since 2009 and marks 16 years of operation. Its homepage FAQ publishes typical investment ranges: an MVP at about $50,000 to $100,000, medium-scale solutions at $100,000 to $250,000 and large or enterprise builds at $250,000 to $500,000. The same FAQ says clients retain full ownership of IP including design and source code, and get complete source code ownership on completion. It names ERIN, Plunge and The Healthy Mummy as case studies.

Not verifiable:the user counts and referral totals attached to those case studies, and the 1,200+ brands figure, are the company's own numbers. It also names SOC 2 Type II and several ISO certifications; I saw the claims, not the reports. Agoura Hills is a real drive from the Westside, so ask where your team would actually sit.

Visit TechAhead

2

Goji Labs

downtown address, broad named portfolio

Score: 9 of 12. Suits: startups and media or consumer brands that want a product studio with strategy and design in the same room.

Verifiable: Goji Labs describes itself as a digital product agency based in Los Angeles and New York, with 800 Wilshire Boulevard in its structured data and a 213 phone number. It says it has 12+ years in business and lists iOS and Android in its stack. It names a long client roster, including KCRW, Fitness Pact, Castability, WWF and Root Insurance. A Clutch snippet shows a $25,000 minimum project size and a $100 to $149 hourly band.

Not verifiable: no pricing or IP terms on its own site. The homepage counters (500+ products, 25 million users, $1 billion raised by clients) are its own claims, and funds raised by clients say more about the clients than about the build.

Visit Goji Labs

3

Dogtown Media

Venice studio with a health tech lean

Score: 8 of 12. Suits: healthcare, fintech and AI app projects where compliance language like HIPAA and software as a medical device matters from day one.

Verifiable: its contact page lists 228 Main Street #4 in Venice. Its about page says it was co-founded in 2011. The homepage names native iOS and Android development and shows named portfolio apps including a Udemy instructor messaging app, a United Nations ESCAP app and a University of Oklahoma innovation app.

Not verifiable:no pricing and no IP terms on the pages checked. The "200+ apps" figure is the company's own. The about page also mentions a presence in San Francisco, New York and London, so as with any multi-office firm, ask which office staffs your build.

Visit Dogtown Media

4

Bee Techy

the cheapest published MVP on the list

Score: 7 of 12. Suits: first-time founders who want a fixed, small, published starting point before committing to a full build.

Verifiable: Bee Techy lists 1601 Vine Street in Los Angeles and names native iOS, Android and cross-platform engineering. It publishes three one-time prices: a $1,000 Snapshot (scope and requirements document), a $10,000 Sprint (clickable prototype and refined requirements) and a $30,000 90 Day MVP described as production-ready with three core workflows and one integration.

Not verifiable:no founding year on its own site, and no IP terms. The named work (TravelPal, Halcyon, Mortgage Maker and others) is real names, but the pages don't always say whether each was a mobile app or a web platform. Three workflows and one integration is a tight MVP; price out what the fourth workflow costs before you sign.

Visit Bee Techy

5

ISBX

Sepulveda Boulevard, enterprise and robotics apps

Score: 6 of 12. Suits: enterprise and brand clients who want a Westside team with named work in clinical trials and robotics.

Verifiable: its contact page lists 3415 S. Sepulveda Boulevard #1250, Los Angeles. It names Science 37 (tagged iOS, Android and CMS), OPIO and BrainCorp as case studies, and shows logos for brands like Red Bull and Sony Pictures.

Not verifiable: no published pricing (the form asks for your total project budget instead), no IP terms, and a founding year only on third-party profiles. Client logos are not case studies; ask which named apps the proposed team personally built.

Visit ISBX

The bottom half isn't worse at building. It publishes less, and in a couple of cases publishes the things larger buyers look for while skipping the things founders look for.

6

Distillery

South Bay mailbox, nearshore engineering

Score: 6 of 12. Suits: companies that want a US-managed firm with engineering capacity in similar time zones, and AI features alongside the app.

Verifiable: its contact page lists a headquarters at 1230 Rosecrans Avenue in Manhattan Beach, with a PMB number. Its about page says it was founded in 2012. It names iOS and Android app development services and lists clients including Tala, Rakuten and Cruisebound.

Not verifiable:a PMB is a mailbox, so I can't tell from the site whether there's a staffed LA office; that's why it scores 1 on office. No pricing or IP terms. Its homepage says it is an Anthropic services partner and an OpenAI partner; I didn't verify either in a vendor directory for this article.

Visit Distillery

7

SPINX Digital

two decades in LA, website first

Score: 6 of 12. Suits: organizations that need a serious website and a companion app from one local vendor.

Verifiable: SPINX Digital says it has been in Los Angeles since 2004, and its contact page lists 911 W. Washington Boulevard. It lists Apple iOS and Android among its technologies and mobile app development among its services.

Not verifiable:its named case studies on the homepage are websites (the LAPD site, Quinn Emanuel, The California Endowment), so it scores zero on named app work. That doesn't mean it hasn't built apps. It means you should ask for app references specifically.

Visit SPINX Digital

8

Sidebench

health-focused product studio, address not shown

Score: 5 of 12. Suits: digital health companies and providers that want a studio with a long healthcare track record.

Verifiable: Sidebench shows a 310 phone number and claims 14 years and 60+ healthcare implementations. It names the nOCD treatment app, plus Baby Steps LA and Cortica as platforms.

Not verifiable:I didn't find a street address on the pages checked, so it scores 1 on office. Third-party profiles place it in Santa Monica or Los Angeles, but they disagree with each other. Its site shows a 4.9 Clutch rating with 48+ reviews; I couldn't confirm that on Clutch itself, so it's not scored. No pricing or IP terms.

Visit Sidebench

9

Swenson He

Beverly Hills address, consumer brands

Score: 5 of 12. Suits: consumer and entertainment brands that want a smaller studio with a Beverly Hills meeting room.

Verifiable: its homepage lists 407 N Maple Drive in Beverly Hills and a Dallas office. Its about page says it was founded in 2014 by two MIT alumni. It names brands including Elf on the Shelf, Gacha World and DropOffer.

Not verifiable:the pages I checked didn't name iOS and Android as capabilities, or say which named brands were app builds; a showcase page may. No pricing or IP terms. Its site says clients give it perfect scores on Clutch and GoodFirms without a count, so that isn't scored.

Visit Swenson He

10

Rootstrap

big bench, thin public detail on mobile

Score: 4 of 12. Suits: larger product teams that want a big design and engineering bench with Latin American delivery hubs.

Verifiable: Rootstraplists Los Angeles as a hub alongside Montevideo, Buenos Aires and Medellín. It says it has 15+ years building production software and names MasterClass, Madison Reed, Tony Robbins and Emeritus.

Not verifiable:no LA street address on the pages checked, and the homepage didn't name iOS or Android work. It says SOC 2 certified; I didn't see the report. A score of 4 is the streetlight effect at work: this is a firm that says little publicly about mobile, which isn't the same as having nothing to show.

Visit Rootstrap

The rank matters less than the fit. A 5 that matches your constraint beats an 11 that doesn't.

Treat the scoreboard as a filter, not a verdict. Decide your one non-negotiable first (a budget ceiling, a written IP position, a healthcare track record, a team you can meet in person) and see who survives it. Then compare the survivors on everything else.

Two rows deserve a longer word, because they're where I see founders make the most expensive mistakes.

The budget-first founder. Suppose you have about $40,000 and a clear idea. Bee Techy's published $30,000 MVP fits, with about $10,000 left over, which is roughly 1.3x its price. TechAhead's MVP floor of $50,000 doesn't fit: your budget is about 0.8x of it. That's a useful answer before any call. It tells you whether you're buying a production MVP, a clickable prototype, or a discovery phase first. If you're not sure which, the cost structure is laid out in what a discovery phase costs in LA.

The "local means onshore" assumption. Several firms here pair an LA address with engineering elsewhere. That can be a perfectly good model and it's often cheaper. But the address on the contact page doesn't tell you where your code gets written. If the time zone and the people matter to you, ask. The trade-offs, with the arithmetic, are in Los Angeles vs offshore app development cost.

If you want the whole city's picture rather than one ranking, our Los Angeles app and AI development hub collects the local guides in one place.

A worked example makes the filter concrete. This one is a scenario, not a client.

Consider a Culver City fitness studio owner with three locations. She wants a booking and membership app on iOS and Android, tied to her existing payment processor, with push reminders. She has about $60,000 approved for the first version. Her lawyer has told her one thing: she has to own the code, because she might sell the business in three years.

Step one is the non-negotiable. The lawyer's requirement is it. Only one ranked firm publishes a source code ownership statement, TechAhead. Everyone else scores zero on IP terms, which doesn't disqualify them, it just means she has to ask. So the first cut isn't a cut at all; it's an email to every candidate asking for the IP clause from their standard agreement.

Step two is the budget sanity check. Her $60,000 sits about 1.2x above TechAhead's published MVP floor of $50,000 and at 0.6x its ceiling of $100,000. It's 2x Bee Techy's published MVP. So she can afford a real first version at either, with very different scopes. Bee Techy's MVP includes one integration; her payment processor would use it up, and push notifications may be extra.

Step three is proximity. She wants to meet in person, because she'll be explaining a class schedule that has more exceptions than rules. Venice, the Sepulveda corridor and Beverly Hills are all a short drive from Culver City. Agoura Hills is not. That doesn't cut anyone, but it changes how she runs discovery.

Step four is the four emails from the list at the end of this article: a fixed-price phased proposal, the named people who'll write the code, the IP clause, and what happens in the 30 days after launch. How fast and how completely a firm answers is the best free signal she'll get about how it runs a project.

Notice what the rank did in all this: not much. The top-scoring firm is the only one that already meets her lawyer's requirement on paper, which is useful. But three firms ranked lower are a shorter drive, and any of them could put the same clause in a contract tomorrow. The total is a summary; the columns are the tool.

The downside of this approach is that it rewards firms that publish, and some very good builders publish very little. If your favorite got a zero somewhere, ask. One email can turn a zero into a two, and the ranking would be the first thing I'd update.

Several figures came up again and again while I built this list, and none of them should travel as fact.

The first group is agency counters. Apps launched, products shipped, users reached, funding raised by clients, portfolio value growth, brands served. I saw all of these on the ten sites, in round numbers with a plus sign after them. Every one is published by the company about its own work, with no method, no denominator and nobody outside the firm checking. They're not in the scores or the table, and where I mention one in a profile, I say whose number it is.

The second group is review counts and ratings I couldn't see at the source. One firm's own blog quoted a Clutch rating and review count from years ago; a search result title for the same firm showed a different count. Another firm's site shows a rating with a "48+" review count. Clutch blocks automated reading, so I couldn't confirm either on Clutch itself. Not scored, and not printed as current.

The third group is AI-written summaries of review sentiment, such as a search summary saying about 80% of one firm's reviewers praised a particular trait. There's no way to trace how that percentage was calculated. Not printed.

The fourth is the "average app costs X" figure. Many LA ranking pages lead with one. In every case I traced, the figure was published by an agency selling app development, with no sample or method behind it. I use only prices companies publish for their own work, labeled as theirs.

If a proposal you receive leans on any of these, that tells you how the rest of it was sourced.

Once you have two or three proposals, these are the lines I'd look for first.

• No clause assigning code and IP to you: Nine of ten ranked firms publish nothing on IP. Get a written assignment in the contract, covering source code, designs and store accounts, before the first sprint.
• App store accounts in the agency's name: Your app should ship under your Apple Developer and Google Play accounts. Moving an app between accounts later is slow and sometimes messy.
• Open-ended time and materials with no phases: You want phases with a price and an exit at each, so a failed prototype costs a prototype, not the whole budget.
• A sales team you never see again: Ask for the names of the people who will write the code, and meet them before you sign.
• No post-launch warranty: Apple and Google change their rules and SDKs every year. Ask what's fixed free after launch, and for how long.
• A counter or statistic on slide one: Apps launched and users reached are marketing. Ask for two references from apps like yours instead.
• Cross-platform or native, decided before discovery: The right choice depends on your features and budget. A firm that decides before understanding your app is selling its default.
• A budget that only covers the build: Hosting, third-party services, store fees and maintenance keep costing after launch. A proposal that ignores them isn't finished.

The native vs cross-platform choice deserves one more sentence. On iOS in particular, some features are simply easier native, and some budgets simply can't pay for two codebases. We lay out the native route and its price range on our iOS app development page, and the Android equivalent on the Android app development page. Whatever you pick, the reasoning should be in writing.

Here is what I chased and couldn't establish, so you can weigh the ranking accordingly.

• Build quality for any firm: No neutral benchmark exists and I commissioned no test builds. The ranking measures disclosure, not delivery.
• Whether an address is a staffed office: A street address on a contact page is the company's statement. I didn't visit any of them.
• Clutch counts and ratings: Clutch blocks automated reads, and counts from snippets and company badges conflicted. None is scored.
• Certifications: SOC 2 and ISO claims on these sites are the companies' own. I saw no reports.
• Pages rendered with JavaScript: Some counters and sections load in the browser and didn't reach my reader. Where that affected a score, I say so in the evidence notes.
• Acquisition status: Searches found no 2024 to 2026 acquisitions for the ten, which isn't proof there were none.
• Contract terms: IP, warranty and liability terms usually live in a master services agreement nobody publishes. Absence on a website isn't absence in a contract.

None of this makes the list less useful. It makes it a shortlist you can audit, which is the most a public ranking can fairly claim to be.

Frenchy Digital is my company. That's a conflict of interest, and it's why this pick sits outside the scored table and is labeled as mine.

So let me score it the same way first. LA office: 1. We're a Los Angeles agency with a 424 number, but I'd be stretching to call what's on our service pages a staffed street-address office, so I'm scoring it the way I scored Sidebench and Rootstrap. Pricing: 2, because we publish price bands on our service pages. Named app work: 2, because our case studies name app builds. IP terms: 2, because full source code and IP ownership transfer is stated. Native iOS and Android: 2, with a service page for each. Years operating: 2, because we have been operating since 2016 in France, with our US company since 2019, which is ten years.

That's 11 of 12, tied with TechAhead. We match on named app work and on years operating, so the tie falls to the last rule, alphabetical order, and that puts us first of eleven. A tie settled by the alphabet is not a win, and I'd rather print that than rig the rubric.

So why is it still my pick?Because the rubric measures disclosure across the whole market, and a particular buyer cares about a particular subset. If you're a founder or operator who wants senior people on the work, a price before the first call, and to own everything at the end, the attributes that matter to you are exactly the two where most of this list scores zero: price and IP.

Here's what that looks like concretely. We're a senior-led, Black-owned app and AI agency in Los Angeles. Discovery and an audit run $9k to $22k over 2 to 4 weeks. On our MVP development page we publish three tiers: $15,000 to $25,000 for a validation prototype, $30,000 to $50,000, and $55,000 to $75,000 and up. Native iOS apps run $50,000 to $250,000 and up. Senior time is $150 to $225 an hour, and retainers run $2,500 to $9,500 a month. You get a fixed-price phased proposal within 5 business days, a 30-day post-launch warranty, and full source code and IP ownership.

Do the same arithmetic I did for the others. Our validation prototype tier starts at about 0.3x TechAhead's published MVP floor, and our middle MVP tier ($30,000 to $50,000) overlaps Bee Techy's $30,000 MVP at the bottom and reaches TechAhead's floor at the top. That's the honest positioning: between the cheapest published MVP on this list and the most established firm's range.

Now the proof, and the caveat that goes with it. Our three Los Angeles case studies are website and e-commerce builds, not native mobile apps. I'm citing them because they show how we scope, build and hand over for LA businesses, not as app outcomes.

For LA Pro Security, we built a site with an instant quote calculator that prices by event type, duration and guard count, a guard scheduling and dispatch dashboard, and a client portal. For Eternal Brilliance LA, we hand-wrote a WooCommerce theme on TailPress and Tailwind 3.4, with ACF Pro product media, AJAX favorites, WooPayments and Square, a separate checkout child theme, and a self-managed CloudPanel server with LiteSpeed. That case study says plainly that no analytics or sales figures were provided, so I have no outcome number to give you, and I won't invent one.

And for Janvier LA, a Shopify store for luxury engagement rings and custom jewelry with customization features and appointment booking, the owner put it this way:

They're very accommodating to our business schedule. More importantly, they make things work.

CEO & Owner, Janvier LA

Our app builds are in the full case study index, alongside the LA work.

You can get from this page to a defensible shortlist in about a week.Here's the order I'd do it in.

1. 1.Write down your one non-negotiable (a budget ceiling, code ownership, a healthcare track record, meeting in person) and use the buyer-type table to cut the list to two or three firms. Open each one's contact page and confirm the address yourself.
2. 2.Email each of them the same four requests: a fixed-price phased proposal, the names of the people who will write the code, the IP assignment clause from their standard agreement, and what their post-launch warranty covers. Note who answers all four.
3. 3.On the call, ask whose Apple and Google accounts the app will ship under, and whether they'd recommend native or cross-platform for your feature list and why. Screenshot every page you relied on, dated, and keep it with the proposal.

Whoever answers those four emails cleanly and those two questions with reasons rather than defaults is probably your developer. Time to send the emails.