Five axes are enough. More criteria do not make a grid more accurate, they only dilute the weight of what genuinely matters.
The five
1. The work. Not whether it is beautiful, but whether it solves a problem that resembles yours. Ask about the assignment behind the work: what was the question, what was the budget, what happened.
2. The team you actually get. Names, not job titles. Who is on your account day to day, how many other clients that person carries, and were they in the pitch room. This is the axis that most often decides between a good and a bad relationship.
3. The fit. Sector experience where the rules drive the work, and ways of working everywhere. An agency used to weekly check-ins does not suit a client who decides quarterly, and the other way round.
4. The price per unit of work. Not the rate, but how much work you get for your budget and which profile does it. See what a normal hourly rate is.
5. The stability of the business. Four of the five numbers you need for that are free in the filed annual accounts. See how to know whether an agency is financially healthy.
Weight them before you meet anyone
Put the weight per axis on paper before the first presentation starts, and have everyone who decides score separately before anyone talks. Do it afterwards and the weight always shifts towards the axis where your favourite happened to be strongest.
What does not belong on the list
- Size on its own. It does not predict quality; the rank correlation between headcount and growth is 0.003.
- The number of awards, unless creative reputation is a goal in itself for your brand.
- Presentation skill. It gets measured anyway, and it already weighs far more heavily than anyone admits.
Drawing up the grid and guarding it is the part where an outsider makes the most difference, because they have no favourite. See Agency selection & pitches.
See also: how to judge pitch presentations fairly and how to compare two agencies.