Arcane Marketing logoArcaneApply
← All posts

Month six arrives and the tone of the meeting changes. The agency brings a deck showing a falling cost per lead and a fuller top of funnel. The owner brings a bank statement and one question: where is the revenue. Both sides are looking at real numbers. Neither of them agreed in writing what the number was supposed to be. Ask why marketing agencies fail and you tend to get soft answers about communication and chemistry. The published record points at something less flattering and more fixable, which is how the engagement was built before anyone ran an ad.

Why marketing agencies fail: the contract shape sets the lifespan

Start with the uncomfortable part for anyone writing this headline, including me. None of the sources cited in this post measures a six-month failure rate. The strongest public dataset points the other way: ANA and the 4A's report that average client-agency tenure now stands at approximately seven years, more than double the 3.2-year average reported in 2016. The same report splits it by model, with integrated full-service agencies averaging 87 months (7.3 years) and media-only agencies averaging 44 months (3.7 years).

The finding that should hold your attention is about review cycles. Clients without mandatory review periods, 60% of respondents, have significantly longer relationships at 8.1 years, against as low as 3.8 years for clients running frequent reviews. Read that as a design result rather than a scoreboard. An engagement built to be re-pitched starts behaving like one. The review calendar trains both sides: the agency spends its energy defending the last quarter, the client keeps a shortlist warm, and the compounding work that needs four quarters to pay never gets started.

One caveat before you carry these numbers into a partner meeting. The samples are enterprise. ANA surveyed the largest US advertisers, and the founder-led equivalent is not publicly measured. Use the enterprise data as directional evidence about how agency relationships respond to contract shape, and treat the six-month pattern you see in your own network as observation rather than a statistic.

Everyone agrees it runs on value, and almost nobody defines it

Buyers are consistent about what they care about. 90 percent of clients say the overall value and long-term ROI of advertising campaigns outweigh cost considerations, and 56 percent cite trust as most important for a successful long-term relationship. The same study then finds the hole underneath all of it: only 10 percent of agencies and 5 percent of clients have a formal, corporate-backed definition of value.

That is the month-six argument scheduled in advance. When value was never written down, each side fills the gap with whatever it can measure most easily, and those two things are rarely the same. One party reaches for delivery metrics that moved. The other reaches for cash collected. Both are being honest. Only one of them is holding the invoice.

Forrester sizes the distance between what buyers want and what reaches them: 80% of marketing leaders say clear communication is critical while only 55% are satisfied, a significant 25-point gap, and seventy percent of respondents prioritize value over cost, yet just 53% feel they're getting it. Treat an agency's communication system as part of the product rather than the wrapper around it. If you cannot tell from the weekly update what changed, what it cost and what happens next, you will be guessing at month six, and guessing owners cancel.

Measurement turns the guess into a fight. Measuring marketing ROI is the number one challenge marketers name for 2026, cited by 33% of respondents, and 37.9% say leadership views marketing as less important to the business than in past years. Budget pressure closes the trap. The percentage of marketers expecting increases in agency investment for digital marketing fell 14 points year-over-year, while the percentage expecting increases for content creation dropped 10 points. Gartner, separately, found that 39% of CMOs are planning reductions to their agency budgets. The agency line now has to defend itself against an in-house alternative every quarter.

Your review window is shorter than the platform's learning window

Here is the mechanical failure sitting under a lot of month-four panic. Google's own documentation says it can take up to around 50 conversion events or 3 conversion cycles for the bid strategy to calibrate to the new objective, and that the learning period is affected when campaigns, ad groups or keywords are added to or removed from the bid strategy.

Run the arithmetic on your own account. Assume an $8,000 monthly budget, an all-in cost of $400 per booked call, and a conversion event that fires only on a booked call. That is 20 events a month. Against a threshold of 50 events, the bid strategy needs roughly two and a half months of spend at that pace before it has enough signal, so a 90-day review lands about two weeks after calibration finished, on a campaign that has barely had a stable fortnight. Halve the budget or double the cost per call and the same review lands mid-learning. The re-runnable version: divide 50 by your monthly count of the event you actually optimize toward, and that is the floor in months for a fair first verdict. It is a model rather than a result, and your account will behave differently. Then add the second-order effect. A nervous mid-flight restructure, the kind that follows an ugly 60-day check-in, affects the learning period again and produces exactly the flat quarter that seemed to justify the panic.

So the contract and the media plan have to agree with each other. If the engagement sets a 90-day judgement window while conversion volume implies 150 days to a fair read, the relationship is scheduled to fail regardless of anybody's skill. Set the judgement date from conversion volume, write it into the agreement, and decide up front which leading indicators you will both accept in the meantime. If your campaign has stalled around $10k a month, check the judgement window against your conversion count before you rebuild anything.

Absorbed scope degrades the work you actually bought

Scope creep gets discussed as a client sin. Agencies own a share of it too. 57% of agencies lose from $1,000 to $5,000 every month on projects and tasks that are unbilled, and more than three-quarters (78%) say they rarely or only sometimes charge for out-of-scope work. That study surveyed 273 managers and executives at agencies across branding, creative, digital, PR, social and web work.

Absorbed scope looks like generosity in month one. By month four it is uncompensated work funded out of the hours that were supposed to go to the core engagement, and what the client experiences is slower turnaround, thinner creative testing and a strategist who has become hard to book. The client never agreed to that trade, because nobody told them a trade was happening. A scope ledger removes most of it: every request logged, priced at zero or priced at something, and visible in the same weekly document as performance. Saying "that is a separate project, here is what it costs and here is what it displaces" in week three is far cheaper than an unexplained slide in month five.

What to put in writing before month one

Define the single success metric and name who owns the data that produces it, including what happens when the CRM and the ad platform disagree. Set the judgement window from conversion volume rather than from the calendar, then put the date in the document. Agree a change-control rule, so structural changes to campaigns require a named decision and an explicit expectation about the reset. Commit the communication artifact itself: same fields, same day, every week, including what went badly. If you are weighing whether the provider should own strategy at all, the difference between a growth partner and an agency mostly comes down to which of those four they will sign for.

The practitioner read: agency relationships rarely end at month six because somebody was incompetent. They end because an undefined success metric met an unmeasurable ROI number in front of an owner whose patience ran out on a schedule nobody set deliberately. Most of that risk can be removed in the first fortnight, before a dollar reaches a platform, by writing down the number, the date it gets judged and the rule for changing the plan. If you want that in place from day one, apply here and bring your last two quarters of numbers.

Sources

Stuck above the $10k ceiling?

We diagnose your account in 24 hours and tell you straight whether it's a creative, audience, or sales problem.

Talk to us