Paid social — Aug 12 to 18

What $1,780 bought in seven days

Paid social influenced $19,489 of revenue and cleared $3,371 of contribution after media — $2.89 back for every dollar spent. Throw away every assisted order and count only the ones where social took the final click, and it still cleared $1,234.

That second number is the floor, and it is the only thing last click is good for. It is the deliberately unfair reading in which every ad that opened a customer's consideration gets zero credit unless it also closed them. Even there, the week clears break-even by 1.7×. The profit is not in question. What is in question is how much of it we are choosing to see.

Everything above the floor is the actual job. Meta is a demand-creation channel — it interrupts people who were not shopping and puts BTO in front of them. Judging it on last click is judging a channel on the one thing it is structurally worst at, and it is the reason prospecting budgets get cut in exactly the weeks they are working hardest.

Four readings of the same week

Paid social touched $19,489. It closed $11,515. The difference — $7,974 — is not a rounding artefact or a measurement error. It is the work Meta is bought to do.

Why last click understates a social channel specifically. Someone scrolling Instagram is not shopping. They see a Bell Moto-10 at 40% off, register it, and come back two days later through a Google search for the model name or by typing btosports.com. Google gets the click; social created the demand. On a last-click scorecard the ad that made the sale happen reads as a wasted impression, and the search campaign that harvested it looks like a genius.

The same logic runs the other way for Meta's own 33.4×. That figure counts anyone who merely saw the ad within a day, and measures against the pre-return value at the pixel. It is the platform marking its own homework and it should not leave the ad account. Any-click is the honest middle: an order Shopify can actually point to, where a paid-social click sits somewhere in the path.

From revenue to contribution — the working number, then the floor

Applying BTO's own measured return rates and cost structure by category, this week's mix clears 27.2%. Run it on the revenue paid social influenced and on the revenue it closed, and both land well above zero. The floor is the proof; the ceiling is the ask.

Now the same walk-down on the floor. Same cost model, same categories, same axis — only the revenue definition is tightened to orders where paid social took the final click. This is the sanity check, not the scorecard.

This is not a new margin assumption. It is the same category-level model already in the profit diagnostic — measured return rates from BTO's own Shopify data, markup held flat at 40%, with freight, processing, return handling and markdown recovery as stated estimates. Applied to this week's actual product mix rather than a catalogue average.

Returns are priced in, not ignored. The clearable rate is applied to gross revenue and already carries each category's own return rate, so the $471 of returns visible in the window is not double-counted — and the returns still to come on these seven days' orders are already accounted for. Returns land weeks after the click; a report showing only booked returns would flatter itself badly.

The honest sentence for the board: "Seven days of paid social cleared between $1,234 and $3,371 of contribution on $1,780 of media. The lower figure assumes every assisted order would have happened anyway."

Where the influenced revenue came from — and what each category clears

Helmets and boots are 67% of the week's influenced revenue — and they are the two categories that clear the least, because they carry the highest return rates in the catalogue.

This is the whole markup-versus-contribution argument showing up in a single week. Helmets influenced $9,254 of gross — half the week — at a 25.6% clearable rate and a 3.91 break-even ROAS. Goggles, gloves and consumables clear 28–32% and break even below 3.30. The blended break-even lands at 3.67 precisely because the mix is helmet-heavy.

Goggles are the quiet good news. They influenced $1,866 across 20 items at a 30.3% clearable rate — the best-clearing meaningful category in the week, and up sharply on the four-day read. Every dollar the mix moves from helmets toward goggles, apparel and consumables raises contribution without raising spend.

Note where the assist layer is heaviest. Pants went from $1,502 closed to $2,364 influenced and helmets from $5,481 to $9,254 — on a last-click scorecard, roughly $3,800 of helmet revenue social touched simply disappears.

Spend by creative type against what it returned

Over seven days, every creative type earned roughly its share of the budget. The catalogue took 46% of spend for 45% of value; UGC took 31% for 29%. UGC's apparent efficiency edge in the four-day cut does not survive the full week.

The honest read: per dollar, UGC returned 0.96× the catalogue's rate over seven days. The four-day snapshot put it at 1.7×, which was a three-day artefact. This matters because it changes the recommendation — the case for more UGC is no longer an efficiency case, it is a capacity case. The catalogue feeds can only sell to someone already in market and they are the ads whose frequency is climbing. UGC is the only creative type that can be manufactured on demand and the only one that puts someone in market in the first place.

The video read is where the real signal is. UGC_Flash_Sale_Bell_Moto10_40%_FH held a 25.2% three-second play rate and carried 556 viewers to 95% — 11.4% of everyone who started it finished, four times any other asset. UGC_Oakley_Airbrake_Goggles_40% hooked 31.7% and Sale_UGC_Youth_Bell_Moto9_FH 31.1%. Those three are the templates to reproduce; the rest lose 97%+ of starters before the end.

Every ad over $3, calibrated to the Shopify basis

Meta's ROAS is scaled down by the account's own 0.33 calibration factor — the ratio of influenced Shopify revenue to Meta-reported value — and set against each category's real break-even. Four ads, $95 of spend, are underwater even on the generous reading.

$95.03 — 5.3% of the week's budget — produced about $33 of contribution against $95 of cost. A net drag of roughly $62 over seven days, about $3,250 a year if left running. Redeployed at the account's own average it would have produced closer to $275, so the real cost of leaving it alone is the $242 gap, not the $62 loss. Materially smaller than the four-day cut suggested — that snapshot flagged $144 across six ads, because two of them had barely spent anything by then.

The whole problem is one ad. Sidi_X-Power_Boots_50%_v1 is $76.87 of the $95. It generated 10,831 impressions and 267 link clicks for two purchases — a 2.47% CTR and a 0.75% click-to-purchase rate against the account's 2.63%. The creative is getting clicks; the page is not closing them.

But it launched on Aug 16, so it has three days of data, not seven. That is not a reason to keep spending — it is the second-largest single-ad spend outside the catalogue and it has produced almost nothing — but it is a reason to check size availability on that SKU before writing off the creative. It is the exact signature of the depleted-size problem already documented in the Channable work.

Ten products, 53% of the week's influenced revenue

Ranked by revenue paid social influenced. Nine of the ten influenced more than they closed, and one of them closed nothing at all on a last-click basis.

Bell Moto-10 Fasthouse Raven leads at $3,404 influenced against $1,166 closed — a +192% assist uplift, the most assisted product of the week, and the subject of the account's best-performing video. Alpinestars Tech 10 LE Air Gold Boots sold $657 with a paid-social click in the path and zero last-click credit — revenue a strict report files under "not social."

The Oakley Airbrake goggles at #2 are the most interesting line in the table. $1,140 influenced across nine units, in the category that clears best in the catalogue — and there are four separate Oakley Airbrake ads running, one of which has the highest hook rate in the account at 31.7%. This is the mix shift the profit work has been arguing for, happening on its own.

Order Protection was attached 46 times across influenced orders — the only order-count signal in this dataset. It is opt-in, so it sets a floor rather than a count: at least 46 distinct orders had a paid-social click somewhere in the path.

The case for more budget — and the honest caveats on it

Efficiency can degrade 65% on the working number, or 41% on the floor, before this week's spend stops clearing. That is real room, and it is tighter than the four-day snapshot implied.

• Change what we report on, permanently. Any-click becomes the working number for paid social; last click travels beside it as the floor, labelled as such. Reporting the floor as the headline has been costing this channel budget in the weeks it was working hardest — $2,137 of contribution was invisible this week alone. One column added to the weekly report ends the argument for good.
• Cut the $95, but check the Sidi SKU first. Four ads are underwater on every measure and $77 of that is one three-day-old boot ad. Pause the three dead ones today; on the Sidi, pull size availability before deciding whether the problem is the creative or the shelf.
• Step spend up 30–40%, not 50%. Current run rate is $254/day, about $7,700 a month, against $29,129 in December — a fraction of proven capacity. But the seven-day headroom is 41% on the floor rather than the 58% the short window suggested, so the increment should be smaller and the learning cycle a full fortnight, watching cost per purchase rather than ROAS, because CPA moves first.
• Fund UGC on capacity grounds, not efficiency. Over the full week UGC returned 0.96× the catalogue's rate per dollar — it is not the efficiency lever the four-day cut implied. It is the only lever that scales, because ASC_Gear_Catalog is already at 2.76 frequency and will fatigue before the audience runs out. Reproduce the three assets with hook rates above 25%.
• Lean into goggles and apparel deliberately. Goggles clear 30.3% against helmets' 25.6% and broke through to $1,866 influenced this week largely on their own. A mix shift there raises contribution on identical spend, and it is the same argument the profit diagnostic makes about the catalogue.
• Give the six mid-window launches a clean week before judging them. Three are showing the best calibrated returns in the account on two to three days of data. That is encouraging, not conclusive.
• Do not extrapolate a flash sale to the quarter. Seven days, 171 influenced items, an active promotion, and August is not December. The stable forecast inputs are the contribution rate and the break-even. The 10.9× is not one of them — as the gap between this report and the four-day cut demonstrates.