How I turned a site from 45% down to 155% up year on year

Between 2020 and 2022 I ran SEO at Digital Next for two brands owned by the same company: Heatstrip, which sells gas and electric heating into residential and commercial spaces, and CrossRay, which sells infrared BBQs.

Same team, same period, opposite seasonal curves. One of them I inherited in decline.

Before any of the numbers, a word about where they come from, because it changes how much weight they can carry.

Four reports, out of a bin

Nothing survives from that engagement except four monthly client reports per brand: January, February, April and May 2022. No third-party rank history exists to fall back on either. The tools that would normally let you reconstruct a domain’s history four years later hold nothing for these two before 2026.

I found the reports in a deleted folder on a backup drive, alongside thirty-two more for other clients from the same period, including the two that document a marine dealership launch. Had that bin been emptied, two years of work would be undocumented.

There is a lesson in that which has nothing to do with SEO: your agency’s reporting tool is not your archive. It is most of the reason I now run client reporting through something I own. When the contract ends, the login goes, and what you can prove about your own career is whatever you personally kept.

Two things about reading them:

Every comparison is year on year. That matters more here than usual, because both brands are seasonal and they peak in opposite halves of the year. A month-on-month reading of either would be close to meaningless.

The direction arrows are vector graphics. Extract the text and every percentage arrives without its sign. I read one as a 20% rise before rendering the page and finding it was a 20% fall. If you are ever mining old PDF reports for a portfolio, render them.

Heatstrip: reversing a 45% decline

Organic search, year on year, from the four surviving reports:

Report	Users, rolling 3 months	Year on year	Monthly users	Goal completions
January 2022	7,789	−45%	1,977	344
February 2022	6,748	−45%	2,633	462
April 2022	11,205	+43%	5,527	796
May 2022	17,237	+155%	8,952	1,927

Monthly organic users went 1,977 to 8,952. Goal completions went 344 to 1,927.

The rows worth looking at are the two −45% ones at the top. Growing an account that is already growing is a pleasant way to spend a quarter. Turning one that is shedding nearly half its organic audience year on year takes longer than the reports make it look, because the first month showing a positive number is a long way after the first month of the work.

Alongside it, the site’s Lighthouse SEO score went from 75 in January and February to 92 in April and May, and held there. Failed audits fell from 11 to 7 across the same window.

That is two independent measures moving together: a technical audit score and an audience. I would not lean on either alone.

CrossRay: revenue, not rankings

CrossRay was ecommerce, so the reporting went further down the funnel. Organic search only, year on year:

Report	Transactions	Year on year	Organic item revenue, year on year
January 2022	94	+91%	+185%
February 2022	64	+36%	+112%
April 2022	48	+108%	+90%
May 2022	67	+24%	+145%

Organic revenue up in every month against the same month a year earlier, by between 90% and 185%.

I am giving percentages rather than dollar figures there for a reason. The reports carry two different absolute item-revenue numbers on the same page, one in the summary card and one in the device-split chart, and in April they disagree by nearly $10,000. I do not know which definition each uses, and four years on there is nobody to ask, so I quote the change rather than a total I cannot reconcile.

On the rankings side it held first position for “infrared bbq” and “crossray bbq” across the window, and second for “electric bbq” at 5,400 monthly searches. Thirty tracked keywords, which is a small set, deliberately: this is a niche product category and the commercial terms are countable on two hands.

Two things a reader should know about the set. March is missing from both brands, and from every other client in that folder, which is a pity because March is where Heatstrip turns from −45% to +43%. And CrossRay’s mid-range keywords drifted while its head terms held: the largest tracked bucket moved from positions 11 to 20 out to 21 to 50 across the window. Both facts sit in the same four reports as the good numbers.

These two brands were also part of a book of sixteen accounts I ran concurrently, which is the normal shape of agency work and worth stating.

What held four years later

CrossRay still ranks first for “crossray bbq” and second for “infrared bbq” today. Four years and other people’s maintenance sit between May 2022 and now, so that is not a current result. What it does show is that the head positions were structural rather than a spike, and structural positions are the ones worth building.

Nothing else here is externally verifiable at this distance. The rank-history tools hold nothing for either domain before 2026 and the SERP archives have no snapshots for these keywords, so the client reports are the record, which is why this piece opens with where they came from.

Why every figure here is year on year

Heatstrip sells heating and CrossRay sells BBQs, so their demand curves are almost exactly out of phase. Read either month-on-month and the number describes the Australian weather rather than the work. Compare CrossRay in January against CrossRay in May and you have discovered that summer ends.

On seasonal categories year on year is the only comparison that means anything, which is why the basis is stated on every figure above rather than left for a reader to assume.