Below target since the week of May 25 after 12 straight weeks at or above goal. June is pacing -7%.
What happened: full-fare customers (new customers paying full price) went below goal the week of May 25. June is ~7% short.
Why: we cut social in April, leaving Google as the only growth channel - and it's saturated. We're spending 54% more on Google than last June for fewer customers, the next customer from generic search costs ~$500 against a $67 target, competitors are leaning in instead of fading, and our own three sites are bidding against each other.
What we do: stop buying the gap on Google, restart Meta before July (the biggest month of the year), and put each site back in its own lane. Demand is healthy and the site converts fine - this is a channel-mix problem, and it's fixable.
10-17% above goal from March to mid-May. Then 99.7% the week of May 25, 92% the week of June 1, and June 1-9 down 4.5% on the same days last year.
Week of Jun 8 is partial (3 days). Source: daily report, Jun 10.
June 1-7 vs the same week last year, all three accounts: 54% more spend, 5% fewer conversions. Almost all the extra money went into generic - non-brand searches like "esa letter" - and the next customer from there costs about 7x our target.
The averages look fine. The margin doesn't - and the margin is where the June volume would have to come from.
| Jun 1-7 | Spend '25 vs '26 | Conversions | CVR | CPA |
|---|---|---|---|---|
| Generic | $61k to $119k (+94%) | 845 to 960 | 11% to 8% | $73 to $124 |
| Brand | $26k to $25k | 725 to 653 | 13% to 12% | $36 to $39 |
| Total search | $97k to $150k (+54%) | 1,772 to 1,680 (-5%) |
Conversions here are Google-reported. The trend matches full-fare customers; the absolute numbers aren't the same thing.
Generic search is at record highs, but total spend is ~8% under the June plan because social is near zero. The average hides the ~$500 margin - so the CAC tile stays green while volume goes missing.
Term by term, May 1 - Jun 9 vs last year: true waste is only ~$13k. The same terms now cost 35-40% more per customer - and the extra spend lands very differently depending on where it goes:
Last year conversion quality improved into peak season. This year it's going the other way - that's not seasonality.
Impression share = the share of available ad auctions a domain showed in. supportpets.com surged in the exact week conversion broke and is now on "pettable" brand searches too. The rest of the pressure is us.
Same pattern on the other top keywords (on "esa pet", usaservicedogs.org went 19% to 42% across April-June). Last year the auction thinned out into peak season - 12.5 competitor domains per keyword down to 10. This year it's fewer, bigger players all pushing at once. The +42% CPC isn't seasonal.
Same total budget as last June, same $66 blended CAC - we swapped social for deeper search and got fewer customers plus an overheated auction.
June 2025: 13,132 full-fare customers at $66 CAC. June 2026 pace: ~12,760 at the same $66.
Cap generic at a marginal CPA ceiling. Pull loose-ESA and registration spend back to last year's depth - roughly $50k/month currently buying ~$400 customers. Keep core and state terms.
Restart at $30-50k with the read agreed upfront: does full-fare volume respond, by device. TikTok waits - when its Q1 ramp got cut 75% in April, brand searches and direct purchases didn't move, so its halo looks weak.
Main owns direct intent; the review sites own comparison and competitor terms, which they convert cheapest. Pull May 15-19 change history on the main and ESA Spon accounts to see whether the expansion was chosen or smart bidding did it alone.
$87k went to comparison queries in six weeks with three of our own accounts bidding each other up (class CPA $64 to $90 YoY). Move them to esaletter.reviews, have pettable.com exit.
July's target is 14,310 - the biggest month of the year. The Meta restart and the generic pullback both need to be live well before then.