Most agency relationships that fail do so inside the first six weeks, and the reasons are usually visible in week one. At The Growth Bully, a Malta performance marketing agency, the pattern is consistent enough to plan around: the work that decides the outcome of a twelve month engagement happens before a single campaign goes live.
So here is the honest timeline. What should happen in the first thirty days, what genuinely cannot be known at that point, what should be measurable by day sixty, and what decision is fair at day ninety. It also covers the part most agency material leaves out, which is what the client has to supply for any of it to work.
If you are still choosing, the groundwork sits in how to brief a marketing agency and the structural question in retainers versus projects. This piece starts the day after you sign.
What should happen in the first 30 days?
Access, measurement and offer work, in that order, with campaigns going live late in the month rather than in week one. Thirty days spent building ads on top of broken tracking produces activity nobody can read, which is the most expensive possible way to begin a retainer.
Access means ad accounts, analytics, tag manager, the website, the CRM and the inbox that receives enquiries, all held by you and granted to the agency rather than created fresh in the agency name. An agency that builds its own accounts over your business is creating a hostage, and that decision deserves more argument than the strategy deck.
Measurement means agreeing what counts as a conversion before anything is counted. A form submission, a phone call over a set length, a booked meeting and a sale are four different events, and an agency that reports all four as leads can show improvement forever. The mechanics of getting this right are set out in conversion tracking that survives, and a diagnostic of the current state is what a marketing audit should cover.
Offer work is the part clients least expect and it moves results most. Before new traffic arrives, somebody has to decide what the visitor is being asked to do, and what happens in the twenty minutes after they do it.
What cannot be known at day 30?
Whether the work is profitable. Thirty days of a considered purchase gives you enquiry volume and cost per enquiry and almost nothing about lead quality, close rate or revenue, because most of those enquiries have not finished deciding yet. Reading month one cost per lead as a verdict is the commonest early mistake.
Platform learning is the other reason to hold back. New campaigns spend the first fortnight finding out who responds, and the numbers from that period describe the search more than the result. Judging creative, audiences or channel mix off it tends to kill the thing that would have worked by week five.
What you can fairly judge at day thirty is professional behaviour. Did the access get granted and used, did tracking get fixed, did work ship when promised, and does the agency tell you about problems before you notice them.
What should be measurable by day 60?
Cost per qualified enquiry, with a definition of qualified that both sides agreed in writing in month one. By day sixty there has been enough volume to separate the enquiries worth a sales conversation from the rest, and the trend in that number is the first honest signal of whether the programme works.
Two supporting measurements belong here. The rate at which enquiries become meetings or quotes, which tells you whether the problem sits in marketing or in sales follow up, and the share of enquiries answered inside the first few minutes, which in most businesses explains more of the gap than any targeting change. The value side of the same question is worked through in what a lead is actually worth.
Day sixty is also when a decent agency starts narrowing. Two or three things should now be getting more budget and attention and several others should have been switched off, with the reasoning stated plainly rather than hidden in a dashboard.
What decision is fair at day 90?
Continue, change the approach, or stop. Ninety days is long enough to see cost per qualified enquiry, early close rates and at least one full sales cycle in most considered purchases, so a continuation decision made on evidence is now reasonable where at day thirty it would have been a coin toss.
The fair version of that decision looks at direction rather than a single month. Three months of improving cost per qualified enquiry with revenue still catching up is a continue. Three months of rising enquiry counts with no change in meetings, quotes or sales is a change of approach, and usually an offer or follow up problem rather than a media one.
Stopping is right when the basics were never delivered. No working measurement at day ninety, no access, no written definition of a lead and no honest account of what failed is a process problem more time will not fix.
What does the client owe the agency?
More than most clients expect, and the engagements that fail almost always fail here first. The agency controls the media and the message, while the client controls the raw material the work depends on, and no amount of competence compensates for missing it.
- Access on day one. Accounts, analytics, website and CRM, granted rather than recreated, with you as the owner on every platform.
- A named decision maker. One person who can approve creative and spend inside a few days. Approval by committee is the quietest way to lose a quarter.
- Honest history. What was tried, what it cost and why it stopped. Hiding a failed campaign only guarantees it gets repeated with your money.
- Fast follow up on enquiries. Marketing cannot outrun a sales process that answers the next morning, which is the whole argument in speed to lead.
- Outcome data back. Which enquiries became customers and what they were worth. Without it the agency is optimising towards form fills and everyone is guessing.
Those five are also a fair test of whether an engagement is worth starting at all. A business that cannot supply them yet should fix that first, or the retainer buys an expensive diagnosis of its own follow up.
Which week one signals predict failure?
Four of them, and each is visible before any money is spent. Campaigns launched before tracking is verified, accounts created in the agency name, no written definition of a qualified lead, and a reporting template agreed after the first report rather than before it.
A fifth is softer and just as reliable. If nobody has asked what a customer is worth to you, the work is being planned without the only number that makes any of the other numbers mean something. Our own version of that diagnostic is the pipeline scorecard, and the structure behind reaching buyers who take months to decide is our Decision Maker Pipeline.
What should reporting look like from month one?
One agreed format, the same every month, arriving on a known date and short enough to read. It should carry spend, qualified enquiries, cost per qualified enquiry, what changed last month and what changes next, with the uncomfortable items at the top rather than buried under impressions.
Agree the template before the first report exists. A format chosen in hindsight tends to flatter whatever happened, and the habit is almost impossible to unwind later. The same discipline sits behind our lead generation programmes.