Ask an agency owner what went wrong with their last white-label specialist and you’ll rarely hear “the campaigns underperformed.” You’ll hear something closer to: the client asked a question in a call and nobody in the room could answer it.
That’s a reporting failure, not a media failure. And it’s the most predictable way these arrangements break, because reporting is the one part of the work that crosses the boundary between the specialist and the client — which means it’s the one part nobody fully owns unless somebody says so out loud.
This is the layer people skip when they set up white-label PPC.
The three models
The agency writes it. The specialist sends numbers and notes; the account manager turns them into the client-facing report in the agency’s template. Maximum brand control, and the client never sees an outside hand. The cost is the telephone game: the specialist writes “we cut spend on the branded campaign because it was absorbing conversions that would have happened anyway,” the account manager writes “optimized branded campaign,” and a real strategic decision becomes a bullet point nobody can defend three weeks later.
The specialist writes it in the agency’s template. The specialist produces the finished client-facing document — same fonts, same logo, same structure the client has seen for two years. The agency reviews before it goes out. This keeps the reasoning intact and still ships under the agency’s brand. It’s the model I’d default to for anything above a small account.
The specialist joins the call. They appear as part of the agency team, with an agency email address if that’s how the agency runs it. Right when the account is technical enough that translating everything through an account manager loses detail — a complex feed, a measurement migration, a multi-market build. Wrong when the agency’s whole positioning is a single relationship owner.
Pick one deliberately. The failure mode is not picking, which produces a fourth model: the specialist writes something, someone edits half of it, and the parts that survive are the parts that were easiest to paraphrase.
What belongs in a monthly report
Most PPC reports are a screenshot of the platform with the agency’s logo on top. The client reads the first number, forms an opinion, and the other nine pages do nothing. A report that earns its place answers four questions in this order:
1. What happened to the business number? Revenue, qualified leads, signups, bookings — the thing the client’s board asks about. Not impressions. Not clicks. If the media didn’t move a business number, say so plainly.
2. What did we decide, and why? Two or three decisions, each with the reasoning. “Moved 30% of budget from Performance Max to Search because search terms showed the PMax spend was going to branded queries” is a decision. “Optimized campaigns” is not.
3. What did we learn that we didn’t know last month? A test result, a segment that behaves differently, a landing page that converts at half the rate of the others. This is the part clients actually remember, and the part that makes renewal conversations easy.
4. What are we doing next, and what do we need from you? A short list with owners. Most stalled accounts are stalled on something the client owes: a feed fix, a page, a conversion definition, an approval.
Everything else — channel tables, device splits, the full campaign list — goes in an appendix or a dashboard the client can open whenever they want. It shouldn’t set the agenda.
The numbers that make reports lie
Three things distort almost every paid media report, and a specialist who doesn’t flag them will eventually get the agency in trouble.
Platform-reported conversions aren’t incremental sales. Google, Meta and TikTok each report conversions using their own attribution, and each is generous with itself. Add them up and you’ll usually have “generated” more revenue than the business actually booked. Report platform numbers as platform numbers, and put the business number next to them.
Widening the attribution window inflates the report, not the results. Most platforms now let you change click and view windows after the fact — ChatGPT Ads shipped flexible 7/14/30-day click windows this month, and the pattern is everywhere. A wider window increases reported attribution. It does not increase sales. If you change the window, say so in the report and restate the prior period on the new basis.
Branded search flatters everything. The campaign with the best ROAS in almost every account is the one charging for people who were already looking for the client. Separate branded from non-branded before you report efficiency, or you’ll be congratulated for a number that means nothing.
More on the attribution side of this in why last-touch attribution breaks and how to calculate ROAS properly.
The access problem underneath all of it
A specialist who gets the ad account but not analytics, not the CRM and not the revenue data can only report what the platform says. They will optimize toward platform-reported conversions, those will drift from real revenue, and the gap surfaces at the quarterly review in front of the client.
Before the first report, the specialist needs read access to GA4 or whatever analytics the client runs, visibility into the conversion definitions, and — for lead gen — some signal about which leads closed. If the client won’t grant that, the agency should know the reporting will be shallower, and should say so in advance rather than discover it later.
A handoff that works
The arrangement I’d recommend to most agencies:
- The specialist sends a draft report plus a short internal note — two or three paragraphs written for the account manager, not the client, covering what they’d say if they were in the room.
- The account manager reviews for tone and client context, and asks about anything they couldn’t defend if challenged.
- The report goes out in the agency’s template, on the agency’s schedule.
- For quarterly reviews, the specialist either joins or writes a longer readout with the reasoning intact.
Step one is the one that’s usually missing. That internal note costs the specialist fifteen minutes and it’s the difference between an account manager who can hold a strategy conversation and one who can only read numbers aloud.
Cadence, and why monthly is often wrong
Monthly reporting is a convention, not a requirement. For an account spending USD 5,000/month with stable campaigns, monthly is fine and weekly is theatre. For an account in a launch, a migration or a test programme, monthly means decisions wait three weeks for no reason.
A reasonable default: a live dashboard the client can open any time, a short written note when something meaningful happens, and one substantial readout per quarter that actually decides something. Related: building a reporting dashboard your team uses and KPI dashboards executives understand.
The short version
Decide who writes the client-facing report before the engagement starts, not after the first one goes out. Put the business number above the platform number. Flag attribution changes in the report that they affect. Give the specialist enough access that the report can be honest. And ask for the internal note — it’s the cheapest thing in the arrangement and it’s what keeps the agency in command of its own client relationship.
If you’re setting this up, the white-label PPC agreement covers who owns the reporting assets when the engagement ends, and white-label PPC pricing covers what reporting should cost you.
Here is how I handle reporting as a white-label partner — including the internal note, which I write for every account whether or not anyone asks.