Selected work

The method behind the madness.

Each of these started with a question the reporting couldn’t answer.

AR/VR software · B2B · US · Global market · $15,000/mo

Finding acquisition in the right market

Objective

Increase SQLs, bring down CPA, and build more B2B business.

Findings

The structure had drifted D2C, optimising toward a consumer-shaped audience that didn’t reflect the actual buyer. The analysis also surfaced an untapped market segment worth planning for.

What was done

Implemented a restructure of targeting and account setup to redirect acquisition to B2B buyers, with the new market opportunity flagged for the forward plan.

Outcome

A four-fold increase in sales-qualified leads within a single month.

4x
Sales-qualified leads
within one month
Machinery parts & equipment · B2B & D2C e-commerce · US · ~$24–35k/mo

It’s not broken, it just needs some organising

Objective

Save the business’s cash flow, through high and low ticket sales, with a tROAS of 3.

Findings

Return on ad spend had been stuck at 1.75x for four months with revenue sliding. Reconciling platform reporting against actual CRM revenue showed the account was optimising toward a signal that didn’t correlate with purchases, products were competing against each other and budget was being spread.

What was done

Products were given a restructure to organise them according to their price points, their categories and what has proven to work. That way we had more control of where the budget would go and what bids were placed.

Outcome

Return more than doubled and lead quality improved materially in the same period.

3.62x
Return on ad spend
was 1.75x
$113k
CRM revenue / month
was $45k
21%
Junk lead rate
was 36%
Commercial & residential luxury tiles · B2B & D2C · US

Recover, recover, recover

Objective

Recover the account after they had experienced website issues that affected their acquisition, while also scaling up.

Findings

The account was still running on pre-incident assumptions: targets it could no longer hit, a definition of qualified demand that had drifted too wide, and campaigns competing against each other for the same searches.

What was done

Reset targets to what the account could realistically deliver, narrowed the definition of qualified demand, and consolidated duplicated activity.

Outcome

Conversion volume nearly tripled at less than half the previous cost.

+184.7%
Conversions
−54.9%
Cost per conversion
+28.7%
Conversion rate
Sunsuits e-commerce · D2C · US · ~$1,700/mo

Consistency is key

Objective

Reach the average industry benchmark and KPI of 2.5x ROAS.

Findings

Creative strategy and frequent creative changes and tests made a big impact in this industry. Competitors were coming up with great designs and changing their ads every 1–2 months, which meant we had to do the same.

What was done

We created a variety of content pieces to help capture the product and its benefits from various messaging angles. We used demand-creation storytelling for our cold audience and demand-capturing storytelling for our warm audience, with new creatives and copy launched every 1–2 months.

Outcome

Our ROAS held consistently above 2.5 and reached up to 3.96x.

3.96x
Peak return on ad spend
KPI was 2.5x
2.5x+
Held consistently
month on month
Commercial & residential luxury pools · D2C & B2C · US · $8,000 (Meta & Google)

Yes, that number is correct

Objective

Capture high-intent demand while building brand awareness and supporting conversions across the full customer journey.

Findings

A lot of the budget was going to people who engaged with the ads but were never going to buy. On the search side, the terms bringing traffic in had drifted away from what someone actually searches when they’re ready to purchase. Cost per acquisition was bouncing around with nothing holding it steady, and the creative had been running long enough that delivery was starting to suffer.

What was done

I cut the segments that looked good on engagement and did nothing for revenue, then rebuilt targeting around people showing real buying signals. New creative went live to fix the fatigue, bid controls went in to steady the cost per acquisition, and I reconnected prospecting and retargeting so the two actually fed each other.

Outcome

We hit a profit of $175,793 and a ROAS of 2,321%.

2,321%
Return on ad spend
$175,793
Profit
from $8,000 spend
Engagement rings · D2C e-commerce · South Africa · R4,500

Engagement buyers

Objective

Bring in leads at low costs.

Findings

We looked into the content and audiences. This led us to testing a variety of messaging and content types across platforms, whilst matching this with our top-performing audiences and what resonates with them. We found that our male audience was not the “I know it all” but more the “help, I don’t know” type, and they responded better to messaging that spoke to them more like a friend lending a hand than a business selling them a product.

What was done

We created new content that spoke to the male audience, in a casual tone of “we got you bro”. We combined fun reels and creative AI for our content, and changed nothing else.

Outcome

Our leads grew by 82.53% in one month, and reached an all-time high of 372 at half the price, decreasing by 45.56%. Quality? So good, they’ve got their hands full.

+82.53%
Leads
in one month
372
Leads
all-time high
−45.56%
Cost per lead
Commercial bespoke furniture · B2B · US

Reaching the right audience, increasing the leads and decreasing the costs

Objective

Increase conversion and lead volume while improving traffic quality and delivery efficiency.

Findings

The targeting was too broad in some places and too narrow in others. A lot of spend was reaching people who were never going to be the buyer, while the audiences that were engaging well were small and hadn’t been built out properly. The creative wasn’t resonating with a senior audience either, and the form let anyone through.

What was done

I cut the audiences that weren’t converting and built out more around the ones that were engaging — same profile, wider pool. Added new creative for that audience, and qualifying questions on the form so the sales team only got real prospects.

Outcome

Lead forms completed at 57.5% against a 35–40% B2B benchmark, with CTR at 1.1%, more than double the LinkedIn average. And it was the right audience doing it: 91% of clicks from relevant job functions, 84% from decision-making seniority.

57.5%
Lead form completion
benchmark 35–40%
1.1%
Click-through rate
2x B2B benchmark
84%
Decision-maker seniority
of all clicks
Geopolitics, economics & business intelligence · B2B · Global

Good things take time

Findings

Campaigns were running as short bursts, around 12 days each. That’s not long enough to get out of the learning phase, so every campaign was still working out who to target right as it got switched off. The conversion tracking was picking up low-value actions too, so the signals going back to LinkedIn were pointing at the wrong outcome. And because each burst started fresh, every campaign began cold — nothing carried over.

What was done

I changed how the bursts transitioned. New campaigns went live before the previous one was paused, so there was always something running and learning rather than a stop-start cycle. Campaign themes moved into the ad layer instead of sitting as separate campaigns, which meant a single campaign could stay live for up to two months and keep being optimised. I also refined the conversion tracking so the platform was working toward actions that actually mattered, and restructured the audiences so seed audiences fed into the right funnel stages — no more starting from zero.

Outcome

2.7x more conversions on 23% less budget, with cost per conversion down 71%. Form and download conversions went from 9 to 175.

2.7x
Conversions
on 23% less budget
−71%
Cost per conversion
175
Form & download conversions
was 9
Industrial electrical equipment · B2B · US · $18,000/mo

Quality over quantity

Findings

The market got more expensive. Competition in the auction had risen, and at the budget the account was running, it could no longer hold presence across a full day — it was buying a shrinking share of a costlier market. The structure underneath compounded it: investment was spread across the whole product category while the demand actually converting had concentrated into a much narrower part of it. And where that demand was strongest, the experience it landed on wasn’t good enough to justify what the click now cost.

What was done

I reset investment to a level the account could sustain across a full day, rather than competing for early presence it couldn’t hold. From there the account was refocused on the part of the market genuinely converting instead of defending the whole category, with weight moved behind proven demand rather than spread evenly across it. I brought the post-click experience up to the standard the traffic now warranted, and built audience intelligence into the highest-value activity so it no longer depended on search terms alone.

Outcome

A 58% quotable rate against Google’s 37% in the same month, with three campaigns at 100%, delivering $952,591 in sales-qualified quotable pipeline.

58%
Quotable lead rate
vs 37% on Google
$952k
Sales-qualified pipeline
in month
3
Campaigns at 100% quotable
Commercial printing · B2B · US · $16,000/mo

Changing with the seasons

Objective

Reach a 4.5 tROAS and recover the account from a decline.

Findings

It wasn’t the structure and it wasn’t the auction. The decline was seasonal — the demand those campaigns had always relied on simply wasn’t there in this period. What had worked for this business every other month of the year stopped working, and no amount of optimising against it was going to bring it back.

What was done

Rather than keep spending into demand that had gone quiet, I researched where the same products were needed next and found a different market coming into season. New campaigns were built around that demand, and investment moved with it.

Outcome

Return recovered from 2.92x to 4.44x, with revenue up 42% to $64,225 on 7% less spend.

4.44x
Return on ad spend
was 2.92x
+42%
Revenue
$45,206 → $64,225
−7%
Spend
month on month

What might your data be hiding?

Free audit