Know Your Numbers

Published: by

Last week, I looked at one very small but important aspect of customer relationships - the human face - in the era of online communications, specifically chat, and how and when you market it. The Chief Marketing Officer of LiveChat, Szymon Klimczak, was kind enough to respond, as well as direct me to a few interesting metrics reports that LiveChat releases every year or two, especially the "Customer Happiness Report."

The Happiness Report is a very well laid out summary of how chat performs for different customers across different categorizations: by industry, by geography, and even across some of the dependent variables. For example, they measure average response time and average customer happiness by geography and industry, but also measure those in reference to each other.

I like this report very much, and credit LiveChat with having the smarts to think about the value of creating it, going to the effort to publish it, and doing so in a visually pleasing manner. Most metrics are, well, mind-numbing. It takes a lot of effort to dig into them and get the valuable data. When I prepare an analysis for a client, it can take me as long to make the presentation concise, interesting and presentable as it does to gather the data and draw conclusions. As my old friend the Rabbi once said, "it takes five minutes to prepare an hour's sermon, and an hour to prepare a five minute sermon." Unfortunately, he said it at the beginning of a sermon for which he had been given five minutes to prepare.

Although I do not have access to the raw data, they appear to have done some good methodology work as well. They excluded customers in trial phase, and grouped together like industries. There isn't much difference between binary options sites and retail foreign exchange sites (if you don't know what those are, you really don't want to know). They also excluded companies who had an average of fewer than 100 chats per month, and performed the analytics across 3,000 companies, with 14MM chats over 3 months. In other words, they wanted the inputs to be statistically significant.

This type of metrics report provides 2 valuable services:

  1. Best practices:  Across the board, if I, as a company with a customer support service, am looking for benchmarks, here they are. Can I compare phone support in a large enterprise software deal with live chat support for a Web service site? Not really. But I can know how good it can get, and might be able to find a close enough industry to draw comparisons.
  2. Leadership: This type of report establishes a thought leadership position for LiveChat. This is not direct marketing. It doesn't say, "buy us because we are the best." It says, "we are thinkers, we are helping the industry across the board, you want to be involved with us in any way possible." Of course, eventually that should lead to market position, market share and sales, but it is more indirect.

The last time I saw a tech company do this really well was the famous mobile app monetization report by Greg Yardley of PinchMedia, since merged with Flurry. Greg used PinchMedia's broad installation of toolkits across mobile apps and devices to measure app usage and ad impressions, correlate them with revenue from ad sales (free apps) and app sales (paid apps), and explain usage and purchase patterns: how price impacts acquisition; how many ad impressions really occur; potential revenue from each, etc. Back in 2009, it really was the first serious analysis of apps as a business on mobile.

Where could the Customer Happiness report go? I found only one serious weakness: it doesn't compare to non-chat.

Sure, it is much harder to get non-chat customer satisfaction and interaction data, but the information is available. This would be invaluable in 2 distinct ways:

  1. Deeper understanding: Why does one industry have a longer response time but higher satisfaction rates? It might be because of accepted standards in the industry; it might be because of the job role of those who interact - some people have more time and patience on a job than others; it might be national cultures. There are many possibilities, most of which likely are known to someone in the industry, but we only can speculate. Correlating all channels - across "multi-channel" - would provide much deeper insight.
  2. Marketing: As a direct benefit to the chat companies and LiveChat in particular, how does chat perform compared to other channels?

The marketing is a key point. The existing report answers the customer question, "How do I do this the best way possible?" Extending it to compare with other channels answers the customer question, "Why should I use this customer communications channel? Give me hard numbers that explain how my world will get better."

In general, though, excellent piece of analysis. I look forward to reading more reports.