3 reasons why we don’t like usual ‘fixed’ KPIs
Most customer service teams use fixed KPIs for their agents: number of tickets handled per day, response time after a ticket is assigned, CustSat score. What these KPIs have in common is that they’re just an abstract number, say 70 tickets per day or 2h response time and that’s why we don’t like them:
1. Fixed KPIs don’t adapt to change
If today is very quiet, there won’t be any tickets for your team to pick up and agents will get unfairly penalised while if your site just crashed and you’re swamped in tickets, you’ll expect them to step up and do more than the basic KPIs. Essentially ‘fixed’ KPIs rely on the assumption that every day will be the same.
2. Fixed KPIs need to be evaluated for every team and situation
Your returns team will not have the same processing times as your first line customer service team, so you’ll need to define different KPIs for each. And to define these, you’ll have to spend time measuring to know what to expect of your agents. Managers spend too much time building up meaningful KPIs and not enough working on the results.
3. Fixed KPIs encourage agents to do only what’s required of them and nothing more.
To alleviate this, managers often include tiers, e.g. 0-30, 30-50, 50-70, 70+. But again, more measurements are required to define these.
Have you tried indexes?
What I call an index is very simple: we measure every agent’s performance against the average of his/her team. For example Jacob is producing 78 tickets a day while his team average is 53. His index is thus 78/53=1.98.
You know Jacob deserves a bonus as he’s 48% more productive than his average colleague, almost as good as one and half agents by himself.
Just like fixed KPIs, indexes can be combined. I used to measure agents on two indexes: productivity on the one hand (volume handled) and effectiveness on the other (customer satisfaction score).
What we like about indexes
- Stop measuring
The main advantage of index KPIs is that there is no need to measure and regularly update your KPIs depending on the team. I’ve used the same index KPIs in Italian and French support teams. As the Italians are compared to the Italians and the French to the French, you don’t need to define different expectations of them
- Adapts to the situation
You just got a new customer service tool (Zendesk?) and it increased your team’s productivity by 20%? No need to review your KPIs to account for this. Since agents are measured against the average of their peers, this one will go up with their individual productivity and there is no change required. Similarly, if you have a very quiet day, everyone’s average will go down and agents won’t be penalised for having handled less tickets.
- Meaningful results
Indexes show you very clearly who is performing well and who is not. Essentially with a score under 1 are doing less than the team average and staff over 1 are doing better. I used to hold an individual monthly review with my agents where we
- Drives effort up
Indexes encourage your team to always work harder. If a good performer joins the team, the average will go up and everyone will have to work harder to stay above the average. This creates a permanent competition to stay above the average, which drives it up and creates a virtuous circle.
Although there are some limitations…
- Critical mass: indexes won’t work effectively in a very small team. You need to have a critical mass of at least 5 agents for them to be effective.
- Harder to understand: agents can find it more difficult to understand indexes than a simple fixed number as they are more abstract. It is very important to spend time explaining the logic of indexes to your agents.
We used to organise a weekly and monthly reward board for the agents with the best indexes every week and month with great success.
Within a large organisation, indexes were used to compare teams to each other and create competition between team managers.