Targets change behaviour whether or not they are sensible, which is why social media kpi planning deserves more care than it usually gets. A badly chosen measure will be met, and the way it is met is frequently the problem.
Choose what the team controls
Reach and follower growth depend heavily on factors nobody on the team decides. Setting targets on them produces either luck or gaming, and often both. Output, response time, completion rate and delivery against plan are all influenced directly by the work.
Ask of every candidate: if this number rises, is that necessarily good? Follower count can rise while the audience becomes less relevant. Comment volume can rise because a post was provocative. A measure that can move for bad reasons is a poor target.
Setting them
- Name the decision or behaviour each measure should drive.
- Prefer measures the team influences directly.
- Define each precisely, including its window.
- Pair any volume measure with a quality one.
- Set the baseline before setting the target.
- Agree how often they are reviewed, and by whom.
- State explicitly what each measure cannot tell you.
Pairing is the most important rule. Volume alone invites more of something regardless of whether it worked; volume beside completion or response quality keeps the incentive honest.
Review the targets, not only the results
A target that is consistently missed may be wrong rather than the team being inadequate. A target consistently exceeded may be too easy, or may be being met in ways nobody intended. Both deserve examination at review rather than automatic renewal.
Keep the set small. Four or five measures can be held in mind and acted on; fifteen become a reporting exercise that nobody uses to decide anything. Dashboard design sits in the reporting dashboard guide, and reporting in the monthly marketing report workflow.
Common questions
How many KPIs should a team have?
Four or five. Beyond that they stop guiding behaviour and start describing it.
Should follower growth be a KPI?
As context rather than a target. Made a target, it pushes toward whatever inflates it fastest.
How often should targets change?
Quarterly at most. Frequently changing targets teach the team that none of them matter.
What if a KPI cannot be measured reliably?
Do not use it. An unreliable measure produces confident decisions on unreliable evidence.
Should clients set the KPIs?
Agree them jointly, and push back on any that the team cannot influence.
What about leading indicators?
Useful, provided you can state the link. An indicator with no argued connection to the outcome is just another number. See contact.
Should KPIs differ per platform?
The definitions often must, since the platforms measure different things. Keep the underlying question the same and let the specific measure vary.