Lesson 8 of 30 · 9 min
Scale versus precision
Every filter you add has a price. Know what you are paying.
Precision is sold at a premium
Tighter targeting costs more per thousand impressions, because you are bidding against everyone else who wants those same people. In ArcPlanner the industry targeting premium is applied on top of base CPMs for exactly this reason — a precise plan buys fewer impressions with the same money.
Do the trade explicitly
Adding a filter is worth it only if the people it removes were genuinely unlikely to buy. A filter that raises CPM by 30% breaks even only if the remaining audience is at least 30% richer in real buyers. Most planners add filters without ever running that comparison.
Small audiences break in a specific way
Under-sized audiences do not fail gracefully. Delivery slows, frequency climbs against the same few people, algorithms lose the volume they need to learn, and results get noisy enough that you cannot tell a good week from a bad one.
Let the algorithm have room
Modern buying platforms optimize best when given a broad pool and a clear conversion signal. Stacking manual filters on top of an optimizing algorithm often removes the very people it would have found. Constrain what is non-negotiable — geography, age gates, brand safety — and let the system explore the rest.
Takeaway
Buy the loosest targeting that still fits the objective, and pay for precision only where it removes real waste.
Check yourself
No score, no signup — pick an answer to see why it's right.
Question 1
A filter halves your addressable audience and raises CPM 30%. When is it worth it?
Question 2
What typically happens when an audience is too small for the budget?