How does personalization affect my CSAT, Handling Times and FCR?

Customer support interactions are becoming increasingly more personalized in both b2b and b2c.

Personalization is important because customers don’t want to be treated as numbers. They want to feel heard and they want personalized interactions. Theoretically, happier customers means higher customer retention, but how does this translate into other business goals?

There is a direct correlation between more personalized customer support and increased revenue. In fact, “customers are typically willing to spend more when they receive such custom-tailored service”. (Forbes)

The challenge, from a customer support professional’s perspective, is as follows:

❓How can I make sure that my team is personalizing each interaction enough?

❓At what point does personalization begin to have diminishing returns (i.e. when do handling times and efficiency suffer with no added benefit in CSAT).

❓ How much does personalizing interactions affect handling times?


Having looked at over 100,000 email interactions, we found some interesting relationships between personalization, CSAT, Handling Times and First Contact Resolution Rate. We want to make it clear that we pride ourselves as data experts, and this sample size is too small to derive accurate conclusions that could apply to most businesses. More research into ticket types, contact reasons, average handling times per industry and more need to be performed. That said, this preliminary article serves as a teaser with interesting findings before we release our in-depth study. 

Personalization Illustration


How does using templates affect your CSAT? 

Customer Satisfaction (CSAT) is the industry standard quality metrics to see how happy your customers are, but it’s being taken more and more as an indicator rather than a fact. CSAT can be hard to measure, and perhaps it’s not representative of your entire customer base. So when we dig a level deeper and discuss the impact of template usage on CSAT, we have to take our findings with a pinch of salt too. 

Nevertheless, we found that interactions handled without templates returned a CSAT 3-5% higher than when templates were used.

But there are some exceptions… Several of our customers actually returned a CSAT scores 3% higher when using templates.

⚠️ Key Takeaway

Customers often know when they’re being greeted with a template response and, since they want to be treated as individuals, they will be (on average) less satisfied.

Healthy template libraries, however, can solve customer’s queries in full, unify your communication tone and increase your customer satisfaction.

How do you know how healthy your template library is?

What is the effect of personalization on CSAT? 

We then looked at what happens when our customer’s modify their templates before sending them out. This means adding a touch of personalization to the template (e.g. filling in additional details to solve the customer query). 

🤷‍♂️ We found that overall, modifying a template does not return significantly higher CSAT scores than unaltered templates. Which means that there is still a gulf between modified/more personalized templates and not using a template. 

We have to be careful here, however, because a high number of the CSAT surveys that are returned may all be from ‘refund requests’ that are accepted and solved with an unaltered template requiring no personalization. So, what this really tells us is that there is no evidence to suggest that personalization has a negative impact on Customer Satisfaction.  

⚠️ Key takeaway

It seems that customers can really sense the difference when a template is used, which could suggest that the templates need more work to please customers. It could also represent that customers are very satisfied by the work of the agent when they come with an atypical request (that can’t be solved with a template). 


How does personalization affect First Contact Resolution (%)? 

A lot.

Unaltered templates will solve only the most basic of all customer queries. Personalized customer service goes beyond just using people’s names and asking them how they are. It means solving customer queries in full, taking into consideration all of the intricacies of what they are saying: “If the agent does not provide a swift, focused, pointed and personalized experience, he creates a point of frustration for the customer”. These customers want their queries resolved first time round, without having to repeat themselves. 

Let’s see how tailoring templates affects FCR (First Contact Resolution %).

✅ When our customers modify templates before sending them, they see as much as a 6% increase in queries answered the first time of asking, as opposed to templates being sent out without personalization. 

But that’s just for unaltered and altered templates. When we compare tickets solved with no template, we see a further increase of between 6-13% FCR than those solved with a template!

⚠️ Key takeaway

Avoiding templates appears to have a very positive impact on FCR. This could be largely down to the fact that when a request is more complex, agents are spending more time to make sure they have fully understood and answered the customer.  

When your team is using templates, therefore, it seems like a no-brainer that they should not be sent out without alterations. A 6%+ FCR could potentially save you hundreds of needless back and forths that put a strain on your support team.   

Does personalization affect Handling Time and efficiency? 

Communication in customer support is a battle between speed and quality. It’s logical to think that writing responses from scratch, or using a template and altering it to fit the customer needs, would increase handling times. We found that to be true. 

✅ From our sample of 100k tickets, we found that Handling Times were, on average, 36% lower when agents used a template.

Going one step further, we also found that adding personalization/modifying templates meant sending out responses 25% more slowly than unaltered responses.

⚠️ Key takeaway

Unsurprisingly, using templates boosts efficiency. That’s not newsworthy! What’s interesting is how significantly (or not) your handling times are affected by template usage. 

Perhaps the question you should be asking is: to what extent can I afford to let my team personalize their interactions so that we maximize efficiency and our qualitative metrics?

There must be a sweet spot, right? 

Nomnom CS Cover Image

Read how Nom Nom perfect personalization at pace with Miuros

What is the ‘right amount’ of personalization? 

The million-dollar question.

Generally speaking, using templates lowers Customer Satisfaction and is detrimental to First Contact Resolution Rate. However, template usage also significantly decreases Handling Time.

But we also know that personalizing templates means higher FCR and, potentially, a higher CSAT. 

So maybe the more accurate question is, when should I personalize customer support and when should I not

For more complex requests, personalization is more important. If you have a look at the templates you're sending out - and the CSAT scores associated with them - you will be able to see where personalization is working wonders and when it’s not having much of an effect. For example, when you’re giving someone a refund, they’ll be satisfied with a generic message as long as they got what they came for - and attempts to over-personalize will likely go unrewarded. 

The question that remains, therefore, is as follows: how can you identify which battles to fight? When a personalized message has low added value for the customer experience, it's usually better to stick to the template and save time, leaving more time for more complex topics that deserve extra care. 

Communication is never uniform and customization shouldn’t be applied at all times. You don't need extreme personalization to reset passwords or handle account closures; part of Communication Analysis means digging deep enough to identify where the opportunities to add real value to your customers (hence, the company) are, and where it's just improving a metric superficially.