The Growth Audit We Run Before Spending a Dollar on Ads

"We're ready to turn on ads. Can you run them for us?"

We get this question a lot, usually from a founder who has just closed a round, or a business leader who has just hired a salesperson, or a CMO who just watched a competitor's ad show up in their own feed. It's a reasonable question. It's also almost never the first question.

Because here's the thing about paid media: it doesn't create demand. It buys attention and points it at something. If that something is ready — consumer product, tech product, what have you — ads work like a multiplier. If it isn't, ads work like a very expensive diagnostic — you'll learn exactly what's broken, one click at a time, at market rate.

So before we spend a dollar of anyone's budget, we run a growth audit. What follows is that audit, in full, as we run it. Take it and run it yourself.

Traffic Is a Stress Test With a Bill Attached

When you turn on paid acquisition, you're not buying customers. You're buying traffic. Traffic arrives at your site, meets your positioning, your landing page, your form, your follow-up, and your pricing — and every weak point in that chain gets found immediately, by strangers, at volume.

Paid media is the fastest way to discover the gaps in a funnel. It's also the most expensive way, by a wide margin, because you pay per discovery.

This audit is the cheap version of the same test. Same gaps, found in a week, with nobody's card on file.

Here's a real example: recently, we were working with a startup team that wanted to grow quickly. Many marketing teams would have turned to paid social and earned media right away. But we decided that we wanted to test the messaging and the product to confirm a product/market fit, and alignment with what prospective customers would actually want.

So we took an organic approach first. We developed anchor content to ground their manifesto, their market positioning, their view on the industry, and to build the founders' reputations. From there, we began having real conversations with prospective customers to garner qualitative feedback on whether the product would be received well in the market. Based on those conversations, we created multiple paid campaigns that allowed us to further test whether the product, positioning, and overall prospective customer experience was working.

We quickly assessed that the messaging was resonating, but customers weren't yet signing up. This led us to do a deeper dive into where customers were dropping off. And we quickly ascertained that it wasn't the ads or the emails, it was the website itself, which was trying to speak to all personas at once. This allowed us to rethink the on-site user experience, and to position the product offerings for specific ideal customer personas, then tailor their journeys from paid campaigns, all the way through the customer acquisition and onboarding, followed by retention.

How to Run This Playbook

• Who runs it. One owner who can pull data and get answers — usually a founder, a head of marketing, or the strategist you've brought in. Six checks, each with a named contributor from sales, product, or ops.
• How long it takes. Five to ten working days. Most of the effort is gathering and interviewing, not producing. It runs in parallel with the rest of a discovery phase.
• What you need before you start. Access to your analytics, ad accounts (even dormant ones), CRM, ESP, and website CMS. Your last four quarters of closed-won data. Thirty minutes each with someone in sales and someone in customer success.
• What you end up with. Four artifacts, listed at the end of this playbook — not a slide deck.
• The order matters. The checks run upstream to downstream, and each is a precondition for the next. There's no point auditing landing page copy if nobody is searching for what you sell. Run them in sequence and stop early if you hit a Stop on Check 1 or Check 6.

The Six Checks

Each check is rated on the same three-point scale:

• Ready — 2 points. Nothing blocking. Spend can flow through this stage.
• Fix First — 1 point. A known gap with a known fix. Spend can proceed only at test scale, with the fix in flight.
• Stop — 0 points. Spending here wastes money at a predictable rate.

Check 1 — Demand

The question: Is there existing demand to buy against, or are you trying to manufacture it?

What to Gather

• Search volume for the problem and the category, not just your brand or product name
• Which competitors are bidding on those terms, and for how long they've sustained it
• Your existing inbound: what people typed, searched, or said before they found you
• Win/loss notes on how customers described the problem in their own words

What to Look For

• Sustained competitor spend on a term is one of the more reliable signals a channel converts — nobody keeps paying for clicks that don't
• Demand for the problem matters more than demand for your solution. People search for what hurts.
• If your customers don't yet have language for the problem, you're category-creating, and paid search is the wrong first channel

Rating

• Ready — meaningful search volume on problem terms, competitors sustaining spend
• Fix First — demand exists but is narrow or seasonal; a limited test can size it
• Stop — no demand to capture yet. Spend the first dollars on content, partnerships, and category education, then come back to paid when there's something to capture.

This check is a gate. A Stop here ends the audit. Nothing downstream can compensate.

Owner: Strategy · Time: 1–2 days

Check 2 — Audience & Offer

The question: Do you know who converts, and what specifically they're saying yes to?

What to Gather

• Existing customer personas, plus the date they were last validated
• Closed-won data segmented by industry, size, role, and source
• The current call-to-action on every active acquisition surface
• Thirty minutes with sales: which prospect converts fastest, and why

What to Look For

• The common failure isn't that founders don't know their customer. It's that they know three or four and haven't decided which one to pay for.
• Personas that predate your last significant product change are stale by definition
• "Learn more about our platform" isn't an offer. An offer is the single thing the ad asks someone to do and the reason they'd do it this week.
• The narrower the offer, the cheaper the click and the better the lead — counterintuitive to almost every founder we work with, and true almost every time

Rating

• Ready — one prioritized persona, validated within the last two quarters, and one specific nameable offer
• Fix First — personas exist but are unvalidated, or the offer is generic. Both are fixable in under two weeks.
• Stop — no agreement internally on who the customer is. Fix this before anything else; every downstream check depends on the answer.

Owner: Strategy + Sales · Time: 2–3 days

Check 3 — Message Match

The question: Does the ad's promise survive the click?

What to Gather

• Every planned ad concept, paired with the URL it points to
• The first screen of each destination page, viewed on a phone
• Current form fields on each conversion point
• Page load time on mobile, on a normal connection

What to Look For

• The most common expensive mistake, and it takes ninety seconds to spot: the ad points at the homepage. Your homepage serves investors, candidates, existing customers, the press, and eight kinds of prospect. An ad is a promise to one person about one thing.
• The headline on the destination should repeat the ad's language closely enough that the visitor doesn't have to re-find what they clicked for
• The ask should be visible without scrolling
• A form requesting company size, job title, and phone number before a first conversation is built for your CRM's convenience, not your prospect's

Rating

• Ready — dedicated destination per campaign, headline echoes the ad, ask above the fold, form asks only what's needed to have the next conversation
• Fix First — pages exist but need copy and form work. Usually a one-week fix.
• Stop — campaigns pointed at the homepage with no dedicated pages planned

Owner: Content + Web · Time: 1–2 days

Check 4 — Measurement

The question: Three months from now, will you be able to tell what happened?

What to Gather

• Google Analytics and Google Tag Manager configuration
• The list of conversion events, and proof each one actually fires
• Your UTM convention, and evidence that someone follows it
• A sample lead in the CRM — HubSpot, Salesforce, whatever you run — traced from source through to closed-won

What to Look For

• Conversion events that were configured once and never verified after a site change
• Source data that survives the handoff into the CRM, not just into analytics
• Whether anyone can currently answer "which channel produced our last ten customers" without a manual reconstruction

Rating

• Ready — events fire, UTMs are enforced, and source persists to closed-won
• Fix First — tracking exists with known gaps. Plumbing work, usually under a week.
• Stop — no reliable conversion tracking. Spending now means arguing about results later with no way to settle it.

This is the least interesting check and the one that most often saves the engagement. Without it, three months from now you'll be in a room arguing about whether the campaign worked, with two people holding different numbers and no way to reconcile them. I've been in that room. Nobody wins it.

Owner: Marketing Ops · Time: 1–2 days

Check 5 — Follow-Through

The question: What happens in the hour after someone fills out your form?

What to Gather

• The named owner of inbound lead response, and their committed response window
• What actually happens on nights and weekends
• The current nurture sequence in your ESP, if one exists
• The first-call script or talk track, compared against what the ad promised

What to Look For

• A lead has a short window where they remember why they filled out the form
• Most inbound leads aren't ready to talk yet, and "not yet" is not the same as "no" — so there needs to be something for them
• Whether what sales says on the first call matches what the ad promised two steps earlier
• A lead that sits in a queue over a weekend is one you paid full price for and then let cool

Rating

• Ready — named owner, committed response window including weekends, a real nurture sequence, and message continuity from ad to first call
• Fix First — response is ad hoc or nurture is missing. Both buildable inside a sprint.
• Stop — no owner for inbound response. You'd be buying leads nobody has agreed to answer.

Owner: Sales + Lifecycle · Time: 1 day

Check 6 — The Math

The question: What is a customer worth, and what can you afford to pay for one?

What to Gather

• Average deal size or first-year customer value
• Qualified-lead-to-customer close rate
• Payback period — how long until you've earned back what acquisition cost
• Any existing cost-per-lead benchmark from past spend, however rough

What to Look For

• With those three numbers you can work backward to what a lead can cost and still make sense, which turns everything above into a budget instead of a guess
• Without them, a forty-dollar lead is a triumph or a disaster depending entirely on numbers that have nothing to do with the ad account
• We've watched teams kill campaigns that were working and double down on campaigns that weren't, purely because nobody had done this arithmetic

Rating

• Ready — all three numbers exist and sales agrees with them
• Fix First — estimates exist but aren't validated against closed-won data. Worth a day to firm up.
• Stop — no view of customer value. You can't tell a good result from a bad one, which means you can't decide anything.

This check is a gate. A Stop here means you can spend the money but you can't learn from it.

Owner: Founder + Finance · Time: 1 day

Scoring and the Verdict

Add the six scores. Maximum is 12.

Gates first. A Stop on Check 1 (Demand) or Check 6 (The Math) overrides the total. Demand missing means there's nothing to buy. Math missing means you can't read the result. Either way the answer is the same.

What the Audit Produces

Four artifacts. If you finish and don't have these, you ran a conversation, not an audit.

1. The scorecard. Six ratings, the total, and the verdict, on one page.
2. The gap list. Every Fix First and Stop, with an owner and a date. This is the actual work product — usually more valuable than the verdict.
3. The offer statement. One sentence naming the persona, the offer, and the reason to act now. Everything downstream is written against it.
4. The measurement spec. The conversion events, the UTM convention, and the field in your CRM that carries source through to closed-won.

Sometimes the Audit Says Don't

We should be honest about where this lands sometimes.

More than once, we've finished a growth audit and told a founder that the best use of their next dollar isn't ads. It's fixing the message match. It's building the nurture sequence. It's picking one persona and building a real offer for them. That conversation is not the one anyone wanted to have, and it's cost us paid-media scope more than once.

But the alternative is spending someone's money to prove something we could have told them in week one. That's not a trade we're interested in. And in practice, the engagements that start with a hard audit and a delayed launch tend to be the ones still running a year later.

This probably sounds like a lot of work to do before the fun part. It is. The fun part is a lot more fun when the funnel underneath it holds.

You can't buy your way past a leak. You can only buy more water.