Agency growth
How do agencies find and win D2C brands that stay?
Find D2C brands that are already spending, on Meta, marketplaces and in AI answers for their category. Qualify them on spend, fixable gaps and fit. Audit the domain before the first call, open with that evidence, pitch a scoped 90 days with a stated expectation, and keep them by answering why every number moved.
344k impressions in Google's AI results for MoolwanSearch Console · Generative AI features · 25 Jun–23 Sep 2026 · real data
What you'll see in 10 minutes
- A list of brands in a category that are actively running Meta ads
- A prospect audit from one domain: site health, AI search visibility, local rank and live ads
- A pitch deck drafted from that audit, ready for your team to edit
Why does this matter before you send another pitch?
A founder messages your team on a Monday: "ROAS dropped over the weekend. Why?" If the account manager answers with a screenshot and a guess, the brand starts quietly talking to other agencies. In our experience, agencies rarely lose D2C brands because the work was bad. They lose them because nobody could explain what changed.
So the way you win a brand decides whether it stays. Pitch with a credentials deck and a promise, and the brand judges you on results you can't yet control. Pitch with evidence about their own business, and they judge you on whether you can explain things. That is a standard you can meet every month.
What it costs to get this wrong. You spend weeks on proposals for brands with no budget, sign brands whose problems you can't fix, and lose good ones at the first unexplained dip. Each of those costs pitch time, onboarding time and a reference you won't get. The method below is meant to stop that before the first call.
Where do D2C brands worth pitching show up?
Look where brands are already spending money or getting found. A brand that spends has a budget, and a brand that's being found has something worth protecting.
Active advertisers in the Meta Ad Library
Meta publishes the ads running across its platforms. Search a category, such as "bedsheets" or "face serum", filter by country and see which brands are running ads now. You can also see how many variations they run and how long each has been live. A brand with many ads running for weeks is spending steadily. A brand whose ads stopped recently may be unhappy with its results. Meta describes what the library holds on its Ad Library tools page.
Marketplaces
D2C brands in India often sell on Amazon, Flipkart, Myntra or Nykaa as well as their own site. A brand with strong marketplace reviews but a weak own site has a clear gap: it has demand but not the margin it could get from direct sales. Write down the brand name, the website and how many products it has listed.
D2C funding news
A brand that has just raised money is about to hire, spend and change agencies. Funding announcements in business news and startup newsletters give you the brand, the round and often the founder's stated plans. Get in touch within weeks, while the plan is being written.
AI answers for a category
Ask ChatGPT, Perplexity and Google "what are the best [category] brands in India" and note who gets named. Brands that are named have something to defend. Brands that aren't named, but appear on marketplaces and in ads, have a gap you can show them in one screenshot. This channel is new, and it is easy for an agency to miss if nobody checks it.
Local search
Search the category on Google Maps for the brand's city. A D2C brand with a store or studio that ranks poorly in the local pack against nearby rivals is easy to show a gap to. The local pack is also where reviews and replies are visible to everyone.
How do you qualify a prospect before spending time on it?
Finding brands is the easy part. Qualifying them tells you which ones are worth a pitch. Score each prospect against three questions.
| Check | What to look for | Good sign | Walk away if |
|---|---|---|---|
| Signs of spend | Live Meta ads, marketplace listings, recent funding, a team page with marketing hires | Ads running for weeks across several variations | No ads, no listings, no team |
| Gaps you can fix | Site health issues, missing schema, absent from AI answers, weak local rank, tired creatives | A gap you can show with a screenshot or a crawl finding | The only gap is "needs more budget" |
| Fit with your services | Gaps that match what your team already delivers well | The fix is work you've done for other brands | The fix needs skills you'd have to hire for |
| Who decides | Founder or marketing head reachable directly | You can name the person and find them on LinkedIn | Decisions sit with a parent company you can't reach |
| Stage and category | A category you understand, a stage where your fee is affordable | You can name rivals and know their seasons | You'd be learning the category on their money |
Swipe the table sideways to see every column.
A prospect that passes the first three checks is worth an audit. A prospect that fails "gaps you can fix" is the most dangerous: they'll sign, expect growth and leave when budget alone doesn't produce it.
How do you audit a prospect from its domain alone?
You don't need access to their ad accounts to show them something useful. The domain, plus public data, is enough for a first call.
1. Crawl the site
Run a crawl of the main pages. Look for critical and high-severity findings: broken pages, slow templates, missing titles, duplicate content. Count them and pick the two that matter most for revenue, such as broken product pages or a slow checkout path.
2. Check product schema and rich results
Look at whether product pages carry valid schema markup. Product pages without it miss out on rich results in Google and are harder for AI systems to read. Use Google's Rich Results Test on a few product URLs.
3. Check AI search visibility
Run the category questions from the discovery step. Note whether the brand is named, which rivals are named and which sources the answers cite. This is often the most persuasive finding, because the founder can check it on their own phone during the call.
4. Check local rank
If the brand has a physical location, check where it ranks against nearby rivals on Google Maps, and look at its reviews and whether anyone replies to them.
5. Read their live ads
Look at what they're running in the Meta Ad Library: how many creatives, how old, and whether the messages differ by audience. Compare them with two rivals. Many creatives that have been live a long time on a single message can point to creative fatigue. Treat that as a question for the call, not a conclusion.
What this looks like on a real brand
We run this audit on Moolwan, the Bengaluru D2C home décor brand that the ai-dm founders own and run. Its September crawl (ai-dm, 150-page crawl, Sep 2026) scored website health at 70%, with 1 critical and 10 high findings. Search Console flagged 24 of 1,424 product pages in its product snippets report (26 Sep 2026). On Google Maps, Moolwan ranks #2 of 17 nearby rivals with a local health score of 90 (ai-dm, Local presence, Sep 2026). In AI search it is strong: 344k impressions in Google's AI results (Search Console, Generative AI features, 25 Jun–23 Sep 2026). An agency auditing Moolwan from its domain would open with the site findings and the flagged product pages, not with AI search, because that is where the fixable gap is.
How do you run a first call that opens with evidence?
Skip the credentials deck. The brand has seen dozens of them. Open with three findings about their business, show the source for each, and ask questions that only someone close to the account could answer.
The order of the call
- Their findings first. Two or three specific, sourced observations from the audit.
- Your questions. What are they measuring, what changed recently, and what did their last agency fail to explain?
- What you can't see. Say plainly what an outside audit misses: spend, margins, conversion rates, returns.
- What you'd do first. One or two moves tied directly to the findings.
- How you'd report it. Show what they'd see each month: what changed, why, the evidence, what you did next.
That last question matters most. Their answer shows you what will make them stay.
How should you pitch and price the first 90 days?
Founders worry about long contracts with vague outcomes. A scoped first 90 days, built from the audit, gives them a decision they can make this week.
Structure the 90 days around the audit
- Days 1–30: fix and baseline. Connect the accounts, fix the critical findings and record a baseline for every metric you'll report on.
- Days 31–60: first moves. Run the one or two moves from the first call, each with a stated expectation of what it should move and by when.
- Days 61–90: review and decide. Compare expected against actual, explain each gap, and agree the next quarter on that evidence.
Price the scope, not the hope
Set a fixed fee for the 90 days, tied to the work: the fixes, the moves and the reporting. Avoid pricing on a ROAS target you can't control in the first month, before you've seen their margins and history. If they want a performance element, add it after the baseline exists, so both sides agree on the starting point.
Put the reporting promise in the proposal
Say in writing that every monthly review will answer what changed, why, with what evidence, what you did and what you expect next. This promise is what keeps them past 90 days. The pitch deck builder turns an audit into slides with the findings already in place, so your team edits the argument instead of formatting slides.
Why do brands leave agencies, and how do you keep them?
In our experience, brands rarely leave because one month was bad. They leave when a number moves and the agency has no answer to "why did this move?" After two or three of those, the founder decides the agency doesn't understand the business.
Answer the why with evidence, every time
When ROAS drops, the answer should name the likely causes, rank them by how confident you are, and point to the record behind each one: a creative that went live, a rival's new campaign, a change in the auction, a tracking break. Moolwan's own Meta numbers show why this matters. Over 30 days to 26 Sep 2026, Meta delivered 2.13× ROAS on ₹28,064 spend against a 4× benchmark, with 171,489 impressions and 3,895 clicks (ai-dm, Paid ads). That is below benchmark, and we report it that way: a why we're still investigating, with the evidence laid out, not a number we hide or round up.
Watch their rivals, not only their accounts
Many unexplained dips start outside the account. A rival launches a sale or doubles its creatives, and your brand's costs rise. Competitor ad tracking watches rivals' Meta ads so the cause shows up in the review, not after the brand has noticed.
Record what you tried and what you expected
Keep a running record of each decision: what was tried, why, what you expected and what happened. When a founder asks "didn't we try that?", the answer takes seconds. When the expectation was wrong, say so. That honesty is what builds trust over time.
How does ai-dm help agencies find and win brands?
ai-dm is built and run every day on Moolwan, where one person runs nine marketing channels in about an hour a day. The same tools are available to agencies for pitching.
- Prospect discovery finds brands in a category that are actively advertising, so your list starts with brands that are spending.
- Prospect audit runs the domain-only checks from this guide: site and SEO audit, AI search visibility, local rank and live ads, with the source shown for each finding.
- The pitch deck is drafted from that audit for your team to edit and send.
- Competitor ad tracking keeps watching rivals after the brand signs.
ai-dm recommends; your team decides what goes in front of the brand.
What should you do next?
Start with your own agency. Run the free agency AI visibility audit on your own domain and see whether you're named when someone asks AI search for agencies in your city or niche. It's a low-stakes warm-up: you'll learn what the audit shows, how to explain it and which findings land, before you put a prospect's name on one.
Then pick one category you know well. Pull ten active advertisers, qualify them with the table above, audit the best three from their domains and book one call that opens with evidence. Win the why on that call, and keep winning it every month after.
Related pages
More guides
What should a D2C brand review show, and why?
How to run a D2C brand review that leads with what changed and why, handles numbers that don't match, presents a miss honestly and records decisions.
Read the guide →How do you reconcile GA4, Meta and Google Ads numbers honestly?
Why GA4, Meta and Google Ads never match, from attribution windows to consent, and a step-by-step way to compare them and explain the gap to a brand.
Read the guide →
Questions agencies ask
Where do agencies find D2C brands worth pitching?
Look where brands already spend or get found. Check active advertisers in the Meta Ad Library, sellers on marketplaces, brands in D2C funding news, brands named in ChatGPT, Perplexity and Google AI answers for a category, and brands in the local pack on Google Maps.
How do you qualify a D2C prospect before pitching?
Check three things. Are there signs of spend, such as live ads and marketplace listings? Is there a gap you can fix with evidence? Does that gap match a service you already deliver well? A prospect that fails any of the three is a slow sale or a short account.
What can you audit from a brand's domain alone?
Site health and technical SEO, product schema and rich result eligibility, whether the brand appears in AI answers for its category, its local rank against nearby rivals, and the ads it is running on Meta. That is enough to open a first call with specific findings.
How should agencies price the first 90 days with a new D2C brand?
Scope the first 90 days around the gaps found in the audit, with a fixed fee for the fixes and a clear statement of what you expect each one to move. Agree up front how you will report the why behind each change.
Why do D2C brands leave their agencies?
Usually because nobody can answer "why did this move?" when a number changes. Brands stay when every review explains the cause with evidence, records what was tried and checks it against what was expected.
No plan to pick. Billed in USD and INR. A brand's data is deletable on request.