I am currently trying to establish how to hold a team of outbound-only reps accountable for production each day. I don't want to set an arbitrary number, but I do find it increasingly difficult to get an accurate production number.

It first started off as:

Calls Made/Calls Scheduled = Production %

120 Calls Made/140 calls Scheduled = 85% Production

Then I realized this isn't a fair assessment, as it isn't taking into account other factors, such as:

  • Handling Emails from customers
  • After Call Work
  • Scheduled meetings and other offline time
  • Breaks
  • Working with other departments offline (i.e. getting help from out IT or design group)
  • They may not need to call every person I assigned them, like if a business is listed twice with the same phone number.
  • System downtime
  • Calls they didn't have to make (but were scheduled to make)

So then the problem turned into this trainwreck, that I needed to calculate each week, for each person.

Production Hours / Scheduled hours - (email hours + ACW - meetings and offline hours + system issue hours + a guesstimate of the time they spent working with other departments + calls they didn't have to make)

This is nearly impossible to calculate as most of the info isn't fully tracked (we're a small, small business) and if it is tracked, it's across multiple systems.

My questions:

  • Do you see an easier way? Do you have any experience?
  • Is there any software that does this or connects to a third-party CRM?
  • Can you help me keep my sanity?
  • 5
    What is the purpose/goal of those calls? Selling stuff, collecting money, making appointments, ...? Because productivity should maybe not just be measured in calls made vs calls scheduled, but the actual results of those calls.
    – AsheraH
    Jan 9 '20 at 13:47
  • @JoeStrazzere - I schedule calls based on a model I inherited that gave rough estimates of time spent doing these things, but it's the manual calculation needed that is tedious and inaccurate
    – Libby
    Jan 9 '20 at 13:50
  • @AsheraH This is the dream. It's not necessarily selling as it is relationship development, we're not a time of need product. I've devised some other metrics to track success on their calls, but still trying to find an average number to compete over a period of itme.
    – Libby
    Jan 9 '20 at 13:51
  • Why do you consider "average number to complete over a period of time" to be important?
    – dwizum
    Jan 9 '20 at 13:58
  • 1
    We don't have a solid way to track schedule adherence -- and the way the team is structured (varying tasks on and off the phones), it would difficult to maintain with manual scheduling.
    – Libby
    Jan 9 '20 at 14:34

In comments, you're getting asked a lot of questions, and you're being asked to explain why you think certain metrics are (or are not) important. This might seem annoying because you're not being given a simple answer, but the reality is, determining which metrics to measure is often not simple. Companies often make the same types of mistakes when determining metrics:

  • They choose things that are easy to measure. If you have certain data at your fingertips, it's easy to talk yourself into a plausible (but maybe inaccurate) reason why it's important.
  • They choose things that are easy to understand. Something like "calls per hour" doesn't need to be explained. That doesn't inherently mean it's useful though.

Unfortunately, measuring the wrong things can lead to worse performance, not better. Measuring the wrong things, and then holding employees accountable to those wrong things, will at best waste people's time and attention by focusing on something that doesn't matter. At worst, it can hurt performance, by teaching people habits that have poor outcomes.

This is especially relevant in a call center environment, where an agent has a choice to make every time they get a cooperative person on the phone. How much effort should they put in to that phone call? If the conversation verges on meaningless small chat, do they let it ride, or push for a quick resolution? Staff measured on something like calls per hour will quickly learn to end calls as quickly as possible which may result in missed opportunities.

So, how do you decide what to measure? The ideal situation is that your business has a tangible goal you can directly link to employee activity. For instance, in a bank, the team that makes outgoing calls to people who are delinquent on their loans can simply be measured on the dollar value of the payments they collect. However, in other outbound call scenarios, the goal may be much softer - maybe it's relationship based, or based on future business opportunities. Even in those cases though, there may be an opportunity to link value back to an employee. In many outbound call centers that are focused on relationship building (in the hopes of creating future business), a CRM can be set up to capture logs of which agents have spoken with which customers - that way, you can retroactively measure someone's performance by watching for new business from customer's they've talked to.

These are just examples though - the key is, before you try to measure something (and hold people accountable to it), make sure you can tie that thing back to your business's goals.

And if you have nothing direct, and are forced to use an indirect measure, it's a good idea to focus on a way to try to validate or audit the strength of that indirect measure. For instance, if you implement a new program measuring an indirect factor, you can watch key indicators of the business's performance during and after the implementation, normalize to remove factors not related to the implementation, and see what the result was. Then, you can turn that effort into a meta-metric to have a fallback where you can measure the metric going forwards. In call centers where something basic like "calls per hour" is legitimately the only type of measurement you have available, it is critically important to understand performance as you implement, such that you have a guard rail to tell you if your new metric is meaningful or not.

As another follow up thought, consider examining specific "case studies" related to the metric, as an opportunity to learn and spread value. If you have a few star performers who always max their numbers, make sure you take opportunities to (manually, separately from the metric itself) validate their success. This can be another way to check that the metric is accurate. And, if it turns out that your star performers really are your star performers, put effort into understanding why they are star performers. People aren't good because they can do a lot of calls per hour. People are good because of what they do on those calls, how they decide when to wrap up a call, and other decisions they make. Don't just focus on the numbers. Let the numbers help you understand where to look for the stories that actually tell where the value is coming from. To say it a different way: the numbers aren't the important behavior, they're just a measurement of the important behavior. And, most importantly, once you know what the behaviors are, you can teach behaviors, instead of teaching people to hit numbers.

And - finally - if it wasn't obvious from the above, if you can't find a reliable metric, it may actually be better to not have one at all. Staff who are measured on metrics that are clearly not meaningful will (at best) become unhappy drones. Staff who are punished based on quality metrics that you can justify will have no counter argument. But, no one wants to be scolded for something that has no value.

  • 1
    Thanks for this response. I certainly don't find the questions annoying -- these are the same questions I need to ask myself and bring up in discussions with the executives. They're nearly obsessed with calls per hour, but it doesn't provide any value as a metric in this business. We're developing relationships, there are no hard sales (but instead future sales). Current systems limit the ability to track $$$, but when that arrives, it'll simplify this process for sure. You actually streamlined some of my thoughts into a better structure.
    – Libby
    Jan 9 '20 at 14:32
  • Sometimes the key is just finding something you can measure with value, and then just start there and move on. Right now, I have a call center I do metrics for that measures whether or not a new customer has a certain field filled in. There's no way for that field to be entered unless certain conditions are met when that customer is on a call with a rep. The data in the field itself isn't valuable, but knowing it is filled in means we know the rep got to a certain point, and we can show (via unrelated reporting) that getting to that point in the call has an impact on future business.
    – dwizum
    Jan 9 '20 at 14:35
  • We have something similar with another metric I'm proposing. We need to improve our data model a little bit to report on it, but, it is the closest thing we have to measuring value. Just have to convince the other stakeholders that production isn't the way. We've tried it, it just makes them run through calls faster and miss the opportunities.
    – Libby
    Jan 9 '20 at 14:41
  • 1
    The good news is, you can sometimes show that a metric is "bad" once you have a way to measure actual value. Even if the actual value measurement is from another system or not directly linked to the metric, you can sometimes manually pick out the trend between the two.
    – dwizum
    Jan 9 '20 at 14:45
  • 2
    Also - another potential idea for outbound calls to build relationships is to capture unqualified contacts and then qualify them retroactively. For instance, you might give a rep half credit for every customer they talk to, and full credit if or when the customer creates business in the future. We do a version of that now - if a call center rep creates a new debit card for a customer, they get a small credit. But then they get a bigger credit after the card has been used 5 times. This helps filter out reps who just pump out cards without actually focusing on customers who will use them.
    – dwizum
    Jan 9 '20 at 14:48

You're asking about potential metrics in a performance management system. Measuring the right values is certainly important, but "imperfect" metrics are fine to use if the rest of your system is well designed.

Some important things to consider relevant to the metrics you choose and your performance management system overall:

  • The end goal of measuring performance is to continually improve performance of the whole group. What matters for the business is total sales, not the sales of a specific individual. Aggregate metrics may be easier to define and derive meaning from.

  • Towards the goal of improving performance, the metrics you track should enable problem solving, not just be an accountability tool.

    • Measure both inputs (e.g., number of calls made, number of calls answered) and outputs (e.g., sales).
    • Know what a winning score is for each metric (e.g., if you know the total sales target for the year, what is the sales target for each week?).
    • Make metrics visible and keep them up to date. A dry-erase board works great.
    • Use the metrics in your morning huddles to reflect on performance of the previous day and think about how to ensure today will be a success.
  • The metrics should incentivize collaboration and teaching within the team. Measuring individual performance and rewarding the "all-stars" creates competition that destroys performance of the group. The all-stars should be incentivized to teach and help, not continue to be a solo contributor.

  • Accountability happens through relationships, not a scorecard. If you have a team member that you believe isn't contributing, have a 1-on-1 conversation to ask what you can do to help while also setting clear expectations.

With those principles in mind, you could consider a set of metrics such as:


  • Booked sales (weekly, aggregated for the whole group, more is better)
  • Cancellations/returns (weekly, aggregated, less is better)


  • New contacts established (an answered call or email, weekly, aggregated, more is better)
  • Average lead age (time from first contact to first sale, weekly, aggregated, less is better)
  • New contacts initiated (weekly, aggregated, more is better)
  • ...

When considering metrics to use, I find it most useful to consider it from an adversarial game perspective: "If these are the metrics, what would someone do who would want to maximize their performance, strictly with respect to those metrics, and what does that mean for my business?"

For example, let's say your metric is "calls completed"; I am rated based on how many numbers of my list of people I call. Then, as an adversarial game, what I am incentivized to do is to call each number, wait for the phone to pick up, and then immediately hang up the phone and call the next number. I have indeed "called" that number, as required, and if I have 100 numbers per day I estimate I could probably complete my daily work in about 1 hour using this method, and I would expect 100% performance, because that's what's calculated.

Now, this seems stupid: Obviously, that's not the intent of the metric. But that is what you are tracking and that is the result, if I wanted to treat it as an adversarial game. I could create a similar example with basically any metric you care to track: Time at desk (sit at my desk and play Minesweeper), time spent per call (try to small-talk with clients as much as possible and waste time, or perform backend off-call tasks as slowly as possible while the customer waits), time spent performing useful work off calls (perform that work as slowly as possible); literally anything you think of can be gamed.

The real answer is, be involved with your team. Don't be a boss, be a leader. Don't say stuff to them like "if you don't complete 85% of this list of calls then I'm cutting your salary" or whatnot. Encourage them to perform their tasks the the best of their ability, and do not draw hard lines on what they must or must not do. Obviously if you have one team member who is obviously slacking, then you can take action against them, but just because one day you have one person who only makes 50% of their calls, doesn't mean that person should be fired; maybe they had an off day, maybe they didn't sleep all night because they were nursing a headache, maybe they had to deal with some very complicated problem in their calls, you don't know. Take a holistic approach and you'll find things will probably work better.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .