A head of growth pushes Meta spend past $150K a month for the first time, and the next morning an email lands: “Welcome to managed accounts.” A name, a calendar link, an onboarding call. Most operators treat it one of two ways — obligatory small talk to bat away, or a free-consulting line where the platform’s employee will explain what is working. Both readings leave money on the floor.
The rep relationship at Google, Meta, and increasingly TikTok is a managed vendor negotiation with a specific structure, a specific set of available levers, and a specific tension: the rep’s incentives are not the brand’s, but the overlap is real and worth working. The rep gets paid on growing account spend. The brand pays for results. Those goals point the same direction often enough to be useful, and far enough apart to matter. Four areas: what a rep can deliver, what a rep cannot deliver, the cadence that captures the value, and how to carry it across the next handoff.
What a rep can actually do for you
A rep at one of the major platforms can unlock four categories of value the self-serve UI does not expose. In rough order of operator importance:
Spend credits. Most platforms run launch-incentive and category-incentive programs that drop credits into accounts for specific moves — testing a new placement, launching a new campaign type, hitting a tiered spend milestone. The credits are rarely advertised; reps surface the ones their book is eligible for when asked specifically. A brand running $1M+ a year that never asks about active credit programs is leaving a recurring line item unclaimed.
Account-level diagnostic views. Reps can pull reports the self-serve UI does not show: audience overlap across active campaigns, attribution-window comparisons, account-wide creative fatigue signals that the per-campaign view buries. Availability shifts by platform tier and over time — the move is to ask in concrete terms (“can you pull an account-wide audience overlap view across the last 30 days”) rather than browsing a menu.
Beta access. Reps can enroll the brand in early-access tests for new placements, bidding strategies, or measurement features. Most betas are not worth the testing cost — they require campaign restructuring, take weeks to read, and end up rolling into the general product anyway. A rough operator read: one in ten turns into a real durable edge. Take the betas that line up with a test the brand was going to run anyway; pass on the rest.
Escalation paths. This is the highest-value lever and the one operators most often forget to test before they need it. When the account gets flagged, suspended, or stuck behind a policy decision a self-serve account cannot appeal, the rep is the channel into a team that can actually act. A brand losing $50K a day to a wrongful suspension is a different conversation when there is an existing relationship and a documented escalation contact. Test the path on something small before there is a real fire.
What a rep cannot do, no matter the relationship
What a rep can do
- Surface spend credits from launch and category incentive programs
- Pull account-level diagnostics: audience overlap, attribution windows, creative fatigue signals
- Enroll in early-access betas (selective — roughly 1-in-10 worth the testing cost)
- Open escalation paths for flagged or suspended accounts
What a rep cannot do
- Lower your CPMs
- Fix bad creative or override algorithmic auction decisions
- Share another advertiser’s data
- Explain a quiet-quarter performance swing with any back-door view into the auction
- Give a different ROAS read than what you see in the dashboard
No back door into the auction. A rep cannot lower CPMs, fix bad creative, override an algorithmic decision, or share another advertiser’s data. When one hints otherwise, treat it as theater. A rep also cannot explain what is driving a performance swing in a quiet quarter — they will try, but the answer is usually a pattern-match against other accounts on their book, not privileged data.
This matters for budget decisions, which is where operators most often misread the rep’s signal. The platform-attributed ROAS the rep is looking at is the same number the brand sees in the dashboard — they are not reading a privileged version of it. When the rep nudges the brand toward more spend on a campaign showing strong ROAS, that nudge inherits all the same platform-attribution blind spots discussed in why blended CAC misleads you when you scale paid spend. The brand still has to anchor budget decisions in a MER-shaped view that the rep does not have access to. The rep cannot close the platform-vs-incremental gap; the brand has to.
The quarterly cadence that works
A quarterly thirty-minute structured review is enough. The brand drives the agenda: what worked, what did not, three specific asks. The asks should be concrete — a named credit program, a specific beta in or out, a documented escalation contact for the next time policy goes sideways. The brand brings its own context; the rep does not construct the agenda.
The most valuable rep meetings have the rep talking roughly 30% of the time. Skip the standard “platform best practices” deck unless there is a specific signal in the account the brand can act on. Anything more frequent than quarterly gets noisy fast — the rep starts surfacing minor optimization suggestions that fill the slot, and the brand starts treating the meeting as a chore. Signal-to-noise is best when real new context has had time to accumulate.
The handoff problem and rep turnover
Reps turn over at the major platforms roughly every 12 to 18 months — promotion, attrition, account-book reshuffles. The brand that builds around the person loses everything at handoff. The brand that builds around documented commitments carries the leverage across.
What operators see
A brand has a $20K credit balance and two active beta enrollments. The rep who set them up gets promoted. The new rep arrives with no file, no context, and a fresh book of accounts to onboard. In six months, the credit balance has lapsed — unused because the new rep didn’t know to surface it — and the beta enrollment is gone from the account. Same outcome with a documented handoff: new rep gets the one-pager on day one. Credits are confirmed in the intro call. Betas are reinstated in the first quarter review.
Keep a one-page rep relationship doc. Send it to every new rep on the introduction email, in the body of the message, before the first call.
One-page rep relationship doc
- Active credit balances and program names (per platform)
- Pending beta enrollments — feature name, enrollment date, next review point
- Named escalation contacts at the next tier up (not just the rep’s email)
- Agreed quarterly review cadence — date of last review, date of next
The reps who respond well to that doc — who treat it as the starting point rather than something to smooth over — are the ones worth investing in. The reps who want to start fresh are the ones the brand should expect minimum value from until the next turnover.
At $1M+ a year in platform spend, the brand is running a vendor relationship. Treating it like one means owning the documentation, not waiting for the rep to reconstruct it.
The takeaway
The rep relationship is not a perk and not an annoyance. It is a managed vendor channel that pays back in proportion to the structure the brand brings to it. Brands at $1M+ a year that treat it as obligatory small talk leave something on the order of a few percent of spend in unclaimed credits, missed betas, and slower escalations. The structural direction is consistent across brands that have tried both postures.
Build the one-page doc this quarter. Run the next review against it. Treat the rep as a vendor — not a friend, not an adversary — and keep the doc current for the rep after this one.