Every agency has a version of the same photograph: a new buyer, three monitors, forty browser tabs, and the thousand-yard stare of someone two days into an account somebody else built.
Inheriting an ad account is archaeology. The structure was designed by a person who is no longer reachable. The naming convention made sense to them. There are campaigns that exist "for a reason" nobody can name, creatives that have run for a year without anyone defending them, and one ad set whose budget is suspiciously protected. Your job in week one is not to improve any of it. It is simply to understand what you have been handed before you touch a dial.
For years that took two weeks. Last quarter it took me an afternoon, and the difference was not effort.
What the two weeks actually went on
Very little of that fortnight was analysis. It was reconstruction.
You export campaign data as far back as the platform will give you. You screenshot the structure so you can see it all at once. You build a spreadsheet mapping which creative ran where, because the naming will not tell you. You ask the client questions they cannot answer, because the person who knew left in March. You find a tracking discrepancy and lose a day deciding whether it is real. Somewhere near the end, you form a view.
The output of two weeks was not a strategy. It was a rough mental model of an account that had been sitting there, fully documented in its own data, the entire time. Every answer you needed already existed. It just could not be asked a question.
Asking the account instead of rebuilding it
The shift is that an inherited account can now be interrogated directly rather than reassembled by hand.
The first hour of my last handover went like this, inside Hawky's copilot. What has actually driven results in this account over the last year. What do the winning creatives have in common. What was tried and quietly abandoned. Which spend is concentrated in things that are no longer performing. Each answer came back with the rows underneath it, so I was not taking anyone's word for the state of an account I had just been handed, including the software's.
By lunch I had what the fortnight used to produce: a map of what is load-bearing, a list of things running on momentum rather than merit, and three specific questions for the client that were actually worth their time.
The part I did not anticipate was the confidence that came with it. Walking into a handover call able to say "this campaign is carrying 60% of your results and this one has not justified its budget since spring" changes the relationship immediately. You stop being the new agency asking for context and start being the one that already read the account.
What the account cannot tell you
This is where honesty matters, because the data has a hard boundary.
The account can tell you what happened. It cannot tell you why the last team chose it. There is always a reason behind the strange ad set: a founder's preference, a retail partnership, a legal constraint on a claim, a burnt campaign nobody wants named. That knowledge is tribal, it lives in people, and no amount of analysis will surface it. Ask the humans, and ask early.
You also inherit the previous team's measurement, not the truth. If their tracking was broken or their attribution windows were unusual, every answer you get faithfully reflects a flawed record. Audit the plumbing before you trust the story, especially on any conversion number that looks better than the client's own revenue suggests it should.
And a year of history describes the world as it was tested, not as it is. An angle that never appears in the account may be missing because it failed, or because nobody thought of it. Only one of those is useful to know, and the data cannot distinguish them.
A handover checklist worth stealing
Four questions, in this order, on the first afternoon. What is carrying the results. What is spending without earning. What did they try that stopped. What is different about the markets or segments that perform.
Then two questions for the humans, which the data will never answer. What is untouchable and why. What has been promised to someone that is not written down anywhere.
That is a full handover. It used to be a fortnight.
Closing Thoughts: the account was always talking
The thing I keep returning to is that none of this information is new. It was all sitting in the account on day one of every handover I have ever done. The fortnight was never spent discovering it. It was spent building a machine capable of hearing it.
We do not need that machine anymore. Which means the first week of a new client relationship can be spent on the part that was always more valuable and always got squeezed: deciding what to do about what you found.
Top comments (0)