While A/B testing is often used within marketing and business intelligence, it's not something that contact centers have traditionally leveraged. Sometimes referred to as split testing or multivariate testing, A/B testing is basically an experiment with two variants - a (fairly) scientific way of testing and measuring so that you can improve effectiveness.
An example is a landing page, or web page, with two versions. Visitors won't know that they are taking part in the experiment but randomly half will view your control page and the other half will view a new variation (ideally splitting the demographics equally). Over time, with enough data, you can see which one outperforms the other. Of course the criteria for success will need to be defined; typically it's more conversions, which might be sales for an e-commerce website or perhaps signups for trials or demos in a B2B scenario.
You might test different offers, page text, colors, headlines, images, or perhaps layout styles. But don't tweak too many things at once or it will become difficult to understand what exactly has driven the improvement that you later go on to observe. Very often what we think will be a clear winner will actually flop, so it can be a real eye-opener and shows we should never assume to know best.
It's also common practice to split test when it comes to email marketing. Typically you would send the same communication out, but with two different subject lines, to a subset of the data. The remaining recipients then receive the email with the winning subject line. Often this is based on open rates, but perhaps click through rates may be more important and therefore that becomes the basis of measurement.
Why Is A/B Testing Creeping Into the Contact Center and How Is It Being Used?
Contact centers are really looking to step up their game at a time when loyalty is fickle and it's getting harder to sustain customer happiness. Consumers have increasingly high expectations and many brands are finding it hard to measure up . With more channels offered and greater mobility, the goal posts have clearly shifted and the pressure is on to deliver low-effort service and amazing customer experiences.
Customer journey planning/mapping is a hot topic and your usage of A/B testing might be centered around this or aimed at determining how best to take a personalized approach to multichannel communication. Small things that have always been done a certain way can be scrutinized and tested in a constant cycle of measurement and improvement.
There is almost no limit to the types of tests you can undertake. The modern contact center has a huge amount of data at its disposal and, just as in marketing, the slimmest percentage of improvements in efficiency can have profound effects on the bottom line. It makes absolute sense to deploy the same approach to testing every element of a campaign in order to learn and optimize.
Using A/B Testing for Journey Mapping, Personalization and Scripting
You can create intelligent contact strategies by applying testing to the contact flows. For example, if you're getting inbound inquiries coming through your website, create options based on the time and the channel of the original contact in order to define the best follow up strategy for your calls, emails and texts. The point is not to harass the customer but to intuitively pick the best times and channels in order to optimize the chances of making contact. You can design, test and refine these contact strategies to improve results.
If you're unsure how much time to give between an initial email contact and a call back, run two versions of your contact flow simultaneously, each with a different time period between email and call back, to see which gets the best call pick up rate. By testing and improving call flow management, you can see some dramatic results.
Just as A/B testing can be used to improve and optimize the online checkout process – tweaking elements of the steps that we take to improve conversions – it can be used to improve the workflow of inbound call treatments to improve the customer journey and customer satisfaction.
Another area to apply A/B testing is agent scripting in order to help improve your call scripts. Similar to the marketing approach of tweaking copy on a website, adjust the wording in a call script and run two or more versions to see which customers respond to best. The same can apply for the written copy used in emails or chat scripts. The more you test and tweak, the more effective your contact center will be.
You can also use intelligent data lookups to offer personalization and improve the customer experience. Again, you can test and refine how and where to use personalization for best results.