A B2B SaaS matched its ads word-for-word and doubled demo conversion

Representative examplesEvery case study in this library is an illustrative composite of the kind of engagement we deliver — written to show our method and standards, not to name clients.

Where they started

A B2B SaaS vendor sells through an outbound-heavy motion: paid search and sponsored placements feed a sales team that lives on demos. Marketing is two people plus a site vendor whose changes run on tickets. Six ad groups target audiences that share a product but not a problem — enterprise buyers, mid-market operators, agencies, and two verticals — and every one of those clicks landed on the same homepage, written some years ago for whoever happened to be reading.

What it was costing

Every ad group landed on the same generic homepage: enterprise buyers read SMB copy, headline promises didn't match the ad that earned the click, and the sales team complained that demos arrived unqualified and confused.

What they could see

• Enterprise prospects replied to demo confirmations asking whether the tool worked at their scale, a question the page never addressed.
• Reps opened calls by correcting the record about what the product actually does, because the homepage promised something adjacent.
• Marketing could not say which ad group produced which kind of demo, because every click shared one URL.
• Copy fixes waited weeks in the site vendor's queue, so ads and pages drifted apart by the month.
• Two vertical ad groups had been paused after their cost per demo looked terrible — numbers that mixed audiences sharing nothing.

The constraints we worked inside

• Six ad groups with distinct audiences (enterprise, mid-market, agency, verticals) had to stay distinct.
• The SaaS brand's legal claims had to be pre-approved — copy could not improvise.
• Sales wanted firmographics captured before the demo, not after.

What had been tried before

A freelancer had built a dedicated enterprise landing page the year before, outside the site's normal build process.

Nobody owned it; after two rounds of well-meant edits it drifted from legal-approved claims, the sales team stopped trusting it, and the page was quietly retired.

The paid team layered on negative keywords and bid adjustments to squeeze mismatched audiences out of the spend.

Filtering trimmed the budget's waste but conceded whole audiences; it treated the wrong page as weather rather than as the thing actually broken.

What we proposed

Give each ad group its own page, built in a tool the paid team controls, and let dynamic text replacement mirror each ad's exact promise so the page continues the sentence the searcher was reading. Two firmographic fields with sensible defaults qualify gently in the form, so sales receives context without friction. Conversion events fire per ad group, which converts the standing mismatch argument — whose copy, which audience — into a lookup. Legal-approved claims are mapped to audiences up front, so approval happens once per claim, not once per page.

Just as important is what we ruled out, and why:

• Building the six pages in the main CMS — The site vendor's change process ran on multi-week tickets, which would have re-created the exact drift between ad and page that these pages exist to end.
• One page with an audience selector — Buyers do not self-categorize before reading, the headline still showed one story per load, and proof sections could not swap cleanly for each audience.
• Personalizing the existing homepage with a script — The site vendor's contract covered template changes only, so injected personalization would have been fragile, untestable, and outside anyone's support when it broke.

How the work ran

Delivered by the growth pod — strategist over 4 weeks, with working increments reviewed with the client every week.

The stack, and the reasoning

Unbounce

The paid team publishes its own pages — that operating reality mattered more than feature lists; any route through the site vendor would have put a ticket queue between an ad idea and its landing page.

Dynamic text replacement

Six audiences share a skeleton but not a promise; DTR keeps the legal-approved claim intact while the words the searcher just read appear in the headline and the first proof point.

Smart Traffic

Within each ad group several intents arrive mixed; Smart Traffic routes visitors to the variant their behavior fits, instead of asking one message to serve a whole audience.

CRM webhook

The two firmographic answers had to arrive with the demo request, not after it; a webhook writes each submission straight onto the opportunity record so reps never ask what the form already answered.

What went wrong

How we worked together

Cadence

Twice-weekly 25-minute calls with the demand-gen lead while pages were in flight; a shared tracker held page status, legal items, and launch dates.

Client side

The demand-gen lead owned the ad-group map and sign-off; a sales director reviewed the two qualifying form questions; legal handled claims in batches.

Decisions

Claim wording was settled in the per-audience workshop and minuted into the claims register the same day; the demand-gen lead owned message choices, and the sales director arbitrated anything the form's questions touched.

They provided

Keyword files per ad group, the approved-claims document, CRM field access for the webhook, and three demo recordings mined for each audience's own language.

What changed

The headline: visit-to-demo-request conversion across all ad groups, 8 weeks after launch versus 8 weeks before — 4.1% → 9.3%, read from Unbounce plus CRM records. A second check: demos rated 'qualified' by the sales team at +27%.

Demos start with the problem the buyer clicked on instead of a correction. Reps read the firmographic answers before the call and open with the right segment's context, so the first ten minutes stopped being repair work. The quarterly budget argument — which audiences deserve spend — became a lookup rather than a fight, because every ad group's demo quality is visible on its own page. And the paid team launches a new vertical page in a day, inside guardrails legal has already seen.

The result was read from Unbounce plus CRM records against the pre-engagement baseline over the stated window, with a guardrail check on demos rated 'qualified' by the sales team. Where platform-reported numbers and business outcomes differ, this record says which layer it is quoting.

What they own now

• The Unbounce workspace with six live pages, variants, and the shared section library.
• A claims matrix mapping each approved statement to the pages using it.
• Webhook field-mapping documentation, including fallback rules for untagged traffic.
• A cloning checklist for standing up a new ad group's page in a day.
• Per-ad-group conversion dashboards inside the paid platform's reporting.