You have successfully added this page to your briefcase. You can return to this page at any time by viewing your briefcase under your account.
You already have this item in your briefcase. It has not been added.
Jul 01, 2003
I read somewhere that some contact centers resolve 78.3% of calls on the first contact. I'd like to know what the formula is for measuring first call resolution? Thanks. -- Jeannie Torres, Modern Postcard
That's a great question, and it really brings to light how misleading some benchmark numbers can be. It is quite possible that the mathematical average first call resolution rate of a sample size of call centers is 78.3%, but when you dig into the details you find that such a number is meaningless. Through our consulting assignments, we noticed that first call resolution was being measured differently just about everywhere we went. To learn more about this, we sent out a survey in June 2003 regarding key performance indicators in general, with a specific focus on first call resolution. When we asked how each participant calculated first call resolution for five different scenarios that typically happen in call centers, the results spoke for themselves. In no case did the most popular of the five answer choices generate even 50% of the responses. In one case, the most common answer had only 23% of the responses. The full survey results will be included in the September/October special edition of Call Center Management Review, available to ICMI Members of the Strategy and Leadership professional interest area. As it relates to first call resolution, though, the data was very clear - there is no common method for calculation. It's a metric that is valuable when you measure your current performance against your past performance (assuming your calculation stays consistent), but there is no value in comparing your results to those of other centers.
I'm currently working on a project to measure FCR and was wondering why customer surveys were used as the measurement. I would have thought it more accurate to track and count CLI numbers. -- David Earl, AA
I have to concur with the points made in Jay’s excellent response.
Directly to your question of what the formula is for measuring first call resolution, with due consideration to Jay’s caution on the many different variations in use, here is the method that we use:
We pull ALL (not just a random sample of) calls made into the call center for a specific monthly period, noting the date of the call and the calling reason(s). We then take that population and look at a window from 5 min to 72 hours both before and after the call in the period. If there are any other calls in the pulled window (for the like account number) we note those reasons as well. The data is then correlated by account (caller) and call reason. Any account calling back within the above window has the call reasons correlated – where they are in the same grouping the call is assigned a “No” to the “Resolved on first call” question for all of the related calls. If no other calls were made by the account in the window, a “Yes” is assigned to the call(s).
As an example:
A caller contacts us to adjust a late payment fee. Two days later the same person calls us to ask for our mailing address (within the 72 hour window).
We would count this as not resolved on first contact because the associate taking the first call should have confirmed that the customer had our (correct) mailing address, thus avoiding the need for the customer to contact us again. The resulting first call resolution percentage would be 0.00% (as we count BOTH calls as not resolved on first call).
To further illustrate the above calculation, assume that five (5) total calls occur in the period, the two (2) detailed above and three (3) additional calls in the period that did not have subsequent (or prior) calls within the window. Our metric would then yield the following:
3 (resolved on first contact) / 5 (total calls) = 60.00% First Call Resolution.
Of course, what enables the above methodology is a robust data mart which allows for quick retrieval and segmentation of the calling data. Without such tools in place you would have to resort to other methodologies, such as a valid surveying (this of course precludes IVRU surveys) of your callers asking the “First Call Resolved” question.
Finally, and I have been a bit long-winded so I will not go into any detail here, we don’t pay any attention to “industry standard” for this metric (or most for that matter). What is important is that we have a goal of 100% First Call Resolution and we work towards that goal using the above methodology for Analysis and Measurement along with Improve and Control tools from our Six Sigma culture.
I hope this has helped!
Best Regards, -- Rick
I think the approach taken should be to have proactive support to agents while on call to move towards a FCR which is 100%. I think this will bring in better efficiency. -- S. Ranganathan, bk Systems
Exactly how you calculate your FCR rates is very important if you want the academic exercise of comparing your call center against others. However, the real power in measuring FCR is not in determining the rate of resolved calls, it's in finding out what's going on with the calls that you didn't resolve. The measurements that you use need to give you information that you can use internally to figure out why people have to make repeat calls to you, and then correct that issue. Then, using the same FCR metric, you would then ensure that your FCR has improved because you have eliminated the source of one of the problems. Then move on to the next one. -- Rob McDougall, Upstream Works Software
The calculation of calls answered without a repeat / total call volume does not make sense to me. Shouldn't the math read calls answered without a repeat / total original calls? Your opportunity to resolve an issue the first time is based on the count of first time calls, so wouldn't it be an unfair penalty to measure over the total call population?
How do you count FCR for a single inbound call in which multiple issues are presented?
If some issues are resolved but not others, should that call be rated as a successful FCR, a failed FCR, or should it be weighted based on the number of issues resolved versus the number presented (and possibly the relative importance of each)? -- Brenda Bell
FCR can also be improved by the following methods:
1. Identify the major buckets which hit the resolution. i.e OSRI.
2. Identify the bottom agents who are contributing to these buckets.
3. Conduct vitality training, controlled calling, etc.
-- Vinayak Seshadri, Wipro Limited - India
It's not the formula for First Call Resolution that is complicated. You simply take the number of "first-time resolved calls" over the total number of calls. But how you define and track those resolved calls is the tricky part.
There's a good article by Donna Fluss in the February 2008 issue of ICMI's Customer Management Insight about FCR, and I blogged about it and how little things seem to have changed since I was first charged with implementing an FCR measure (without any significant technology to use, either, of course!).
Here's a link to the blog post -- "First Call Resolution: Great Principle, Hard to Apply"
In addition, consider reading the ICMI Member Research Report on First Contact Resolution. It's free for ICMI Members, but anyone can download it here. You can read my comments about it in my blog, too: "First Contact Resolution Research Report"
First call resolution should be measured from both the perspective of the call center and from the customer. Post-call customer surveys will reveal whether the caller perceived that all of their issues have been resolved. It is important not to measure that too quickly since the issue may have been something that takes time to ensure is resolved. For example, if an item was charged to the customer in error, the customer won't know for sure that it has been resolved until the next bill comes out. Measuring only from the call center's perspective removes the customer's point of view from the equation and that misses the point in my opinion. -- Maggie Klenke, The Call Center School
Yankee group published a terrific white paper on the subject that i think you will find valuable, my company sponsored it (in the spirit of full disclosure).
You can google the following to find it - free:
Best Practices for Implementing
a First Contact Resolution
Program in the Contact Center
Their conclusion from the white paper is below:
... The automated analytics-based performance management
software approach to FCR evaluation is clearly the superior
method in terms of accuracy and call coverage measurements.
It involves little if any human intervention and is very much driven by a consistent flow of objectively collected data. -- Ronald Hildebrandt, Enkata
It is really interesting reading the different comments and views on FCR. On the one side we see FCR being pushed everywhere as one of the top KPI’s that we should be measuring in the call centres. On the other hand though there doesn’t seem to be a generic methodology of calculating it. Speak to five different people and you will most probably get five different views.
Didn’t somebody once say “if you can’t measure it then it can’t be important”? So if FCR is so important (and I agree it is) then we need to know how to measure it correctly. So against one of the previous responses – I do think that the formula for First Call Resolution is complicated – because it is so important.
Like many of you I have also reviewed and analysed multiple methods and ways to find the true calculated value of First Call Resolution so that we can use this as a global measure within our company (we were actually looking for First Contact Resolution, but found out that we don’t really know how to calculate First CALL resolution- so we should start there first).
Against some of the previous comments in this thread I do not believe that the FCR calculation should be conducted against the total call volume. That “initial” call from the customer will always be there, the customer will always have a first call – even if you resolve it straight away. So including these calls will give you an unclear measure.
The biggest hurdle to get over is the definition – and definitions that you’ll need to use to identify a call as resolved first time. Just as important is the period in which you want to measure (this will be different for each organisation). For example our reporting period to define FCR is 7 days.
When that “initial” customer call comes in we have no way yet of recognising it as a single call or a repeated call – at this stage it could be both. So the first definition to measuring FCR is to firstly be able to identify and measure REPEAT CALLS.
To understand FCR we need to difference between calls that were resolved first time (a) and calls that we not resolved first time (b) because is it only the number of calls not resolved first time that generate repeat calls. To obtain this is down to your reporting capabilities and how you capture your customer data i.e through CLID or through unique entries in your CRM system – no one can tell you what is the right or wrong way to do this. We simply use a repeated CLID in 7 days (yes I know this has flaws as well, but we have seen it provides the most accurate identification).
Example - we receive 615 “initial” customer calls on Monday (we don’t yet know if the are single or repeat calls). We then measure Monday + 6 days later and identify that we have 150 repeat calls from this initial customer base based on the CLID. At this stage you need to understand that these 150 repeat calls may be from customers that have called more than once in those 7 days. This is important as I have seen that some centres calculate FCR off the back of a repeat call % i.e in this example 150/615 = 24% repeat call and then they say the remainder of 76% must therefore be FCR – WRONG!
Back to definitions – at the moment we have only calculated repeat calls – lets call this (c) in the formula. Based on the CLID we now know the number of unique customers that generated these 150 repeat calls or as mentioned before – the number of calls that were Not Right First Time (b) and say it is 115 customers that generated the 150 repeated calls. (In our company We do not take into account if a customer has called back more than once, but have the view that 1 call back on the same subject is 1 too many and we haven’t done a good job in the initial call)
By identifying (c) and (b) we can now calculate (a) - our calls that were Right First Time or the true FCR number.
We therefore have:
615 initial calls
C = 150 repeat calls
B = 115 not right first time calls
So 615 minus (b) = (a) which is 500
A= 500 right first time or in other words A=FCR
FCR= A/(A+B) and in this example = 81%
It is hard getting people to understand this to start with, but we have found that by measuring and understanding both repeat calls AND FCR gives us greater ability to understand and address the issues that are causing our customers to have to call us more than once and helps us in reducing the need for customers to contact us in the first place (which is a totally different subject!!)
Regards -- Scott
If your performance reports collect transferred calls on inbound calls, you'll get partial but essential data to build a customized formula.
I agree that first-call resolution is an important metric, actually the #l driver of customer satisfaction. So it's most important that you measure it -- what gets measured gets managed. However, there is no standard measurement method. In my opinion, it's imperative that you measure it by your own defintion of what is first-call resolution, and then benchmark against yourself, being consistent in what and when you measure. Furthermore, I'd focus on the second percentage, not the 86% best-practice FCR rate, but the 14% that are repeat calls -- those that cost you in external and internal customers, and that negatively impact your bottom line.
This is a great discussion and the common issues are the differing contact centre call handling environments which make a universally acceptabe or practical, measure impossible.
But what everyone appears to be missing is thast FCR is not the best measure of Customer satisfaction, but it is certainly close.
Given that we are nearly all in multi-channel environments, surely the real aim is to minimise CONTACTS per customer or transaction or sale and so so while FCR goes a long way toward achieving that, there will still be calls & contacts customers have to make, regardless of the quality of the contact centre interaction because of other failures and shortcomings in the business. Why did they have to make the contact in the first place??
So we strongly believe the measure should be "Contacts per x" where "x" is either orders, customers, services, product lines etc, and then we look at FCR after that to really deliver higher customer satisfaction at a much lower cost. We have deployed this approach across many industires for over 15 years with great success.
In summary you can have a great FCR, but still way too many calls and contacts for your customers' liking.
To effectively measure successful resolutions made at the service desk, you need to break down your resolution reporting into three distinct measurements as follows:
- First Contact Resolution rate: this number should represent the total incidents resolved during the first contact made by the end user. This number should only be measured against those incidents that are resolvable by the service desk. True FCR is a measurement of your efficiency and knowledge of the topic that you should be supporting. Today’s service desks respond to a myriad of requests, incidents and problems, many of which should not be included in your FCR calculations since they are only managed escalations and do not truly reflect your ability to resolve incidents locally.
- In Scope Resolution rate: this number represents the total incidents resolved by the service desk versus the total incidents that were “in-scope” or “resolvable” at the service desk. This is similar to the FCR measurement above with the exception that it looks at all resolutions made at the service desk and not just those made on the first contact. A good practice is to break this report down further by displaying the length of time in intervals that it took to resolve the incident reported (FCR, 1-4hrs, 4-8hrs, etc.).
- Remote Resolution rate: this number is the total number of incidents resolved by the service desk versus all incidents reported. This measurement is the “value” of the service desk.
The ultimate goal of continuous improvement is to increase the amount of issues that you can resolve locally at the service desk, thus increasing your Remote Resolution rate and increasing the number of incidents that are “resolvable”. Effective reporting of the top call drivers coupled with a good feedback mechanism from your resolver groups will assist in finding ways add to the “scope” of what your service desk does, and ultimately drive more value to your organization.
Vice President - Strategic Partnerships
Please sign in to contribute an answer. If you don't have an account you can register for free here.
Need help with something specific? Search our entire QueueTips section to find it.
Can't find what you're looking for in a current QueueTips post? Submit a new post to us!
© 2013 UBMAll Rights Reserved