Work that delivered
Tap any project to open the full case study.
Ecommerce Rebuild & Product SEO
A B2C ecommerce catalog that couldn't be found. Product pages are the money pages — and the ones most often left with manufacturer boilerplate, thin copy, and no keyword intent — so the store ranked for its own brand name and little else. This was a discovery problem, not a traffic one: the site was invisible to the people searching for exactly what it sold, and those buyers were checking out somewhere else.
Two workstreams, deliberately separated so each could be measured on its own. The rebuild turned existing traffic into revenue — site structure, product-page layout, and the path to checkout. The SEO program earned new, non-brand traffic through a repeatable template applied across the entire catalog, rather than hand-tuning a handful of hero products. Tracking them as separate outcomes kept the measurement honest.
- New product-page template with keyword mapping behind every page
- Titles and metadata restructured across the full catalog
- Internal linking between related products and categories
- Site-speed improvements and structured data
- Product keywords woven into an ongoing blog program
- Abandoned-cart email flows to recover lost checkouts
Organic Instagram Growth
Instagram is A. Cakes' primary acquisition channel, and by early summer the account had gone flat. It still got views, but almost all of that reach landed on existing followers — the content was circulating inside an audience that had already decided how it felt about the brand. The month before any changes, only 58.8% of views came from non-followers and the account added just 14 net followers. The problem wasn't content quality; it was distribution. For a business where a booking is worth a significant amount and the buying window opens 9–18 months before the wedding, an account that only talks to its existing audience isn't doing its job.
I set the goal as cold reach, not followers — followers are a lagging number; reach is the one you can actually influence. Get the content in front of people who'd never seen the brand, give them a reason to follow, and the follower count takes care of itself. That broke into three decisions: concentrate on the format Instagram was actually distributing (Reels) instead of spreading effort evenly; build a real content calendar to publish consistently and learn what worked; and give new visitors a specific reason to follow — because reach without a follow is just a view.
- Built a monthly content calendar — set the cadence, mapped concepts ahead, and batched production so the account never went quiet in busy weeks
- Shifted the mix toward Reels and produced for discovery — the opening seconds had to land for someone who'd never heard of the brand
- Pulled back on Stories, which only reach existing followers and were absorbing time needed elsewhere
- Reworked what a new visitor sees on the profile so there's a clear reason to follow, not just a nice grid
| Metric | Before | After | Change |
|---|---|---|---|
| Accounts reached | 8,909 | 31,633 | +255% |
| Views | 32,359 | 70,060 | +117% |
| Interactions | 1,202 | 3,369 | +180% |
| Net followers | +14 | +83 | 5.9× |
- The format shift did the work. Reels grew from 13K to 54K views (~4×) while Posts held flat and Stories fell — Reels went from 40% to 77% of all views. The gain came from reallocating effort to one format, not posting more everywhere.
- It broke past the existing audience. Non-follower views rose from 58.8% to 76.6% — cold reach went from ~19,000 to ~53,700 views. That was the actual problem I set out to solve, and the number I'd point to first.
- Engagement rose while reach expanded. Interactions per view went from 3.7% to 4.8% (+30% relative) — the opposite of the usual pattern when an account breaks into cold audiences.
- The profile converted better per person. Follow rate per account reached went from 0.16% to 0.26% (~+67%) — credit to the profile and follow-reason work, measurable separately from the content.
Instagram Insights, straight from the account. Tap either image to view full size.
Bridal Show on a $500 Budget
A bridal show only works if both sides show up — and neither commits first. Vendors won't pay for a booth without a guaranteed audience, and attendees won't register for a thin vendor list, so the campaign has a cold-start problem built in before any marketing happens. It also had to pay for itself and then some: this wasn't awareness for the venue, it was booth revenue plus a pipeline of couples who'd eventually book the space — on $500 in paid media and no team. That budget couldn't buy enough reach to fill a show, so the strategy had to come from somewhere other than spend.
Sell vendors first, and treat every signed vendor as a distribution channel, not just revenue. A vendor with money on the line wants attendance as badly as I do — and each has an audience of engaged couples already following them. So instead of buying reach I couldn't afford, I built the vendor relationship so promoting the show was easy and obviously in their interest, and their audiences became the acquisition channel; the $500 primed the pump and filled gaps. The second decision: treat the attendee list as the real product. Booth revenue pays for the show, but the couples who walk in are venue leads — captured properly, the show pays out twice.
- Vendor revenue — owned outreach, the pitch, and booth sales, then all vendor communication and coordination from signing through show day; kept vendors engaged enough that promoting the show felt like their idea, not an obligation
- Attendee acquisition — ran the registration site and the signup campaign, and built cross-promotional assets so vendors could market the show to their own lists without having to make anything themselves
- Lifecycle communication — owned every attendee touch: registration confirmation, pre-show, and the post-show follow-up that turned attendees into venue leads
| Outcome | Result |
|---|---|
| Vendor booth revenue | $40,000 |
| Venue sales from show leads | ~$50,000 |
| Total revenue attributable | ~$90,000 |
| Return on paid media spend | 180× |
- Marketing generated the revenue, not just spent it. The $40,000 in booth sales came directly out of the campaign work — not leads for someone else to close. That's unusual, and it's the number I'd point to first.
- The show paid out twice. The ~$50,000 in venue sales came from leads captured at an event that had already covered its own costs several times over — the case for treating the attendee list as the actual product, not a nice extra.
- The budget wasn't the constraint I expected. $500 couldn't buy attendance, and the vendor cross-promotion channel likely outperformed what that money could have bought anyway — those audiences were already qualified and already trusted the source. The constraint pushed a better structure than a bigger budget probably would have.