Case Study · B2B SaaS

Put $1 Into Ads.
Get $5.65 Back.
And That’s the Pessimistic Math.

AskBruce is a multiple-five-figure B2B SaaS for spirit-wear team dealers — a founder who won his first ~200 customers in person, one handshake at a time. His ad account was flying blind and nobody trusted it enough to scale it. Thirty days after we took over: 285 leads, 38 qualified demos, subscribers arriving at a quarter of the price the founder said he’d gladly pay — and the honest version of that ratio is 11.29:1. Keep scrolling.

0.00:1

Return per $1 (Pessimistic)

0

Leads in 30 Days

0

Qualified Demos

0%

Lead Cost Drop

0x

Daily Budget Scaled

Ran on
Meta Ads
Facebook
Instagram
01 · The Client

AskBruce

Software for team-dealer shop owners: custom spirit-wear designs, quotes, and orders in minutes instead of days. Around 200 paying customers before we started — a multiple-five-figure monthly recurring revenue business built almost entirely at trade shows. Goal on record: grow from 200 to 1,000 users in a ~15,000-dealer market.

Product

B2B SaaS · Spirit-wear dealers

Goal

Qualified demos → subscribers

This write-up

First ~30 days · Jun 15 → Jul 16, 2026

02 · The Problem

An ad account nobody could see inside

Every customer so far had been won in person, and the founder knew that channel was topping out. Ads were already running — small, cautious, and completely blind.

The trade-show ceiling

“Once they're in my herd, I'm not gonna let him go. But I just don't have a very big herd of people.” The in-person channel that built the business couldn't scale it.

Flying blind

The founder had no way to see what was being done in the ad account or whether any of it worked. Only 10 customers ever had come from ads — and the proof was “they're just telling us they found us on Facebook.”

Optimized for the wrong thing

The account was tuned to get clicks on a button. But a click isn't a customer — the real sale happens days later, on a demo call, where Meta couldn't see it. The algorithm was hunting clickers, not buyers.

The result of all three: nobody — not the founder, not the person running the ads — felt confident putting more money in. You don’t scale a machine you can’t see inside. So the first job wasn’t more budget. It was making every dollar traceable — which ad, which lead, which demo, which subscriber.

03 · The System

Quality over volume, measurement over vibes

Ads went live the same day as kickoff — at 10x the daily budget the account had ever run — after fixing the message first (spirit wear, not uniforms). Then four deliberate choices did the compounding.

A funnel that filters

Ad → landing page → qualifying questions → booked demo with the sales rep. Every layer filters. Fewer demos get booked, but the ones that do show up ready to buy.

Tracking the sale, not the click

Subscriptions close on sales calls, where Meta's pixel is blind. We send every real subscription back to Meta as a conversion event — so the algorithm learns from buyers, not button-clickers. The exact thing the old setup got backwards.

"Never Change the Winner"

137 ads ran. One winner emerged. Its budget is untouchable while challenger ads fight for the remaining share — and challengers must keep earning 20% of the booked meetings to stay alive.

Weekly reviews, zero side-quests

A standing weekly numbers review with the founder. Shiny distractions (Reddit, LinkedIn) deliberately turned down: don't change arenas and don't split focus while the main channel is compounding.

The Funnel

285 leads → 38 demos → 1 in 4 subscribes

Thirty days, measured end to end — every stage traceable back to the ad that started it.

Leads captured
51% cheaper by week 4
285
Qualified demos booked
13.3% of leads
38
Paying subscribers
26.3% of demos become customers
1 in 4

The friction is the feature: qualifying questions before the calendar meant fewer demos — and 1 of every 4 demos becomes a paying customer.

04 · The Timeline

Leads got 51% cheaper while the budget grew 2.7x

The usual trade-off is: spend more, pay more per lead. A winning creative plus clean buyer-signal bent the curve the other way.

Weekly Trend · Jun 15 → Jul 16, 2026

Budget grew 0.0x. Each lead got 0% cheaper.

Weekly budget (x of week one)
Lead cost (% of week one)
Subscribers landed
Hover or tap any week — leads got cheaper while the budget climbed
1x2x3x50%75%100%100%−42%−44%−51%−48%Jun 15Jun 22Jun 29Jul 6Jul 13*

−51%

Lead Cost, Week 1 → 4

2.7x

Budget Scaled

285

Leads in 30 Days

1 in 4

Demos Becomes a Sub

*W5 is a 4-day partial week — budget shown as daily rate vs week one.

  1. Jun 15Kickoff call — ads live the same day
  2. Jun 15Message fixed before launch: spirit wear, not uniforms
  3. Jun 15Qualifying funnel live
  4. Jun 23"One Tab Wins" launches — the account winner
  5. Jun 29First subscribers land
  6. Jul 7Weekly review: every reported number true or understated
  7. Jul 6–12Peak week — 96 leads at the month's cheapest price
  8. Jul 14"Exceeded my expectations" — on record
  9. Jul 14The bottleneck flips: lead gen → sales capacityGood problem
05 · The Creative Engine

204 creatives. One story won.

Wide testing found the signal; discipline protected it. Here's every creative from the first 30 days — and the one that took half the spend.

0

Creatives Produced · The First 30 Days

204 creatives produced in 30 days — 137 launched as ads, and one winner took half the money.

One Tab Wins

An 18-minute reply beat a shop still in Illustrator

Launched June 23. Within a week Meta's own budget allocation was voting for it — and it never stopped climbing.

“One Tab Wins”

The Winner

A coach texts two shops. One replies with a mockup in eighteen minutes. The other is still in Illustrator… Not because the mockup was better. Because it arrived.

A story, not a feature list — paired with a surreal magenta/teal visual that looked deliberately alien next to every normal mockup ad in the feed. It sells one thing: speed of reply.

49.9%

Share of all spend

17% cheaper

Lead cost vs account avg

7.15%

Click-through rate

172

Leads generated

“Done in one tab”

Paused

Same theme. Same product. Same “tab” keyword. Written as a feature claim instead of a story.

Proof that the story wins, not the keyword. A feature-claim version of the exact same idea cost 3x more per lead and got cut.

3.1x

Lead cost vs winner

2.6x

Lead cost vs account avg

12%

Share of spend it earned

Cut

Verdict

Meta voted with the budget

Share of the whole account’s spend that Meta’s own algorithm gave the winner each week — no manual forcing.

Jun 22
26%
Jun 29
51%
Jul 6
73%
Jul 13
75%

Independently confirmed in the CRM: 45% of all 285 leads carry the winner’s tracking tag.

Why not just make more ads?

Because the failure mode isn’t too few creatives — it’s killing a compounding winner out of boredom. “Never Change the Winner” means the winner’s budget is untouchable, and new ideas must prove themselves on merit.

Four weeks in, challengers were still earning 20% of booked meetings — enough to keep the pipeline of future winners alive without cannibalizing the current one.

What the Ads Were Made Of

Every creative that took real spend, sorted by the angle it leads with.

Speed-to-reply stories~37

The winner's DNA

Losing to faster shops~18

"The first shop wins"

Pain callouts~16

Illustrator grind, five tools for one job

Price-first angles~12

Cheap-vs-an-employee framing

Pure urgency~10

Never earned real spend

Go Deeper, Not Sideways

The biggest creative family was 10 variations on one winning visual — digging into what works, not chasing novelty. When something works, you make more of it, not more of everything.

Small Ads That Overdelivered

"Beat the Hours Lost"34% cheaper
"Three Steps. Done."27% cheaper

Lead cost vs the account average. Small-budget challengers kept alive to earn their 20% of booked meetings.

06 · The Math

A customer for a quarter of what he was ready to pay

On the first call, the founder named the highest price he'd still gladly pay for one new customer. Here's where the system actually landed.

The Price of a Customer

0.0x cheaper

On the very first call, the founder named the most he’d gladly pay for one new customer. The system delivers them at less than a quarter of that price.

what we pay
0the price he said he’d happily pay

2.15 mo

Each customer pays for itself

~2 mo

Whole ad budget repaid by new revenue

26.3%

Demo → paying customer

2.7x

Budget scaled meanwhile

The Equation

The absurd ratio, stated honestly

Two ways to run the LTV:CAC math. We report the pessimistic one and show our work on both — flip the toggle.

12 months

of one customer’s revenue

÷

what one customer costs

measured, not guessed

=

0.00:1

We told the client to assume customers stay just 12 months — even though this product’s customers historically stay about two years. The number we report is the pessimistic one. Flip the toggle to see what we actually expect.

3:1 — what SaaS people call healthy
5.65:1 — this account, pessimistic math
11.29:1 — this account, realistic math

2.15 months

Payback Per Customer

Each new subscriber pays back their own acquisition cost in just over two months — everything after that is profit on a customer who stays for years.

~2 months

Whole Budget Repaid

The entire 30-day ad budget is repaid by new-customer revenue in about two months — before counting the 38-demo pipeline and 285 leads still being nurtured.

True or understated

Reporting You Can Audit

Every number reported to the client on weekly calls checked out against raw data as accurate or conservative. 153 leads claimed; 173 actual. That's the direction errors should point.

Within four weeks, the bottleneck moved from getting leads to handling them — seven demos landed on the closer’s calendar in a single day. The founder’s verdict: “a good problem to have.” The next move is a CRM and smarter demo routing, not more spend — you fix the pipeline before you pour more into it.

07 · In His Words

Thirty days in, on record

Both quotes from the week-four review call, word for word.

Exceeded my expectations… it's exactly what I was looking for and wanting. Couldn't be happier.
A

Founder, AskBruce

30 days in

It's growing my business fast, it's finding leads I wasn't able to find before… pouring gasoline onto the fire of building up a successful startup.
A

Founder, AskBruce

30 days in

Want this system pointed at your SaaS?

A funnel that filters, tracking that follows the sale all the way to the subscription, and a creative testing method that protects winners instead of burning them. Same playbook, your product.

Book a Discovery Call

This case study covers the first ~30 days of an ongoing project (June 15 – July 16, 2026). Metrics come from Meta Ads Manager, the CRM, and server-side conversion tracking. The reported 5.65:1 LTV:CAC assumes customers stay just 12 months, on a product whose customers historically stay about 24. Absolute spend and revenue figures are intentionally withheld. Client details and quotes shared with permission.

Pareto 13 Ltd, registered in England & Wales, company no. 15014377. VAT no. GB 525 4202 23.
Registered with the UK Information Commissioner's Office for data protection, no. ZC241334.
© 2026 Pareto 13 Ltd. All rights reserved.