In my earlier columns I have addressed several ways that you can improve customer experience using Big Data analytics, and one of the most successful approaches is applying speech and data analytics to your existing reservoir of customer interactions.
One of my earliest experience with speech analytics was at Amazon.com during the 1999 holiday season. I was running worldwide customer service for Amazon’s three web sites (US, UK, and Germany) and that holiday put a lot of pressure on our operations. Despite major advances in web self-service and operational efficiencies, Amazon launched new “stores” that Fall that required returns (unlike earlier stores for Books, Music, and Video) and greater levels of support, and then the WTO met in Seattle in early December where I had both of the US contact centers in a “no protest zone” that seemed to encourage protesting.
Riots ensued, service levels dropped, and in one of our daily War Room sessions my IT partners offered to apply a new technology called word-spotting to help relieve the email queues. We figured out that all email messages with “Santa Claus”, “Christmas”, or “ruined” merited faster attention so they helped us to create a new email queue assigned to our tier 3 customer service reps with those key words, sometimes all of them such as “If you guys don’t get Johnny’s toy to us before we head out to our parent’s house, Santa Claus won’t arrive and Christmas will be ruined!!!!”
It took some heavy lifting but our queues rapidly got back into control, we saved many Christmases for Johnny and Rachel, and we could track positive customer word of mouth (similar to NPS today) as a result of this intervention program; plus, my agents really liked fixing customer problems, and here they could “lock” or remove multiple email messages that were starting to accumulate since Mom was getting so anxious about getting her products in time. The following year we expanded this analytics program and it’s become much more common since then, but how can you harness its power today?
Collect and act on NPS-powered customer feedback in real time to deliver amazing customer experiences at every brand touchpoint. By closing the customer feedback loop with NPS, you will grow revenue, retain more customers, and evolve your business in the process. Try it free.
Here are five easy steps to take applying speech and data analytics to improve your customer experience:
1. Collect your customer voices
You might want to review my March 2017 column “Using Big Data to Build an Integrated Voice of the Customer Program: A 6-Step Guide”1 where I outlined 20 or more different places where you can collect customer voices, bring them together in an “Integrated VOC program”, and analyze their trends using customer contact reason codes or key words.
For example, you should include social media posts that you control (company Facebook site or Twitter feeds) and that you do not control (e.g., Google, Yelp, and industry sites); your customers’ chat sessions, email, phone contacts, and text messaging; verbatim comments on any surveys (e.g., post-contact web- or IVR-based questionnaires, 3rd-party outbound calls, and website performance sites); and any interactions with your sales teams, executives, field support or repair, tellers, and other customer-facing teams.
2. Catalog key words and hot topics
This is a key step, and can be very revealing. Here I’d suggest pulling together a cross-functional team across the entire customer “journey” with your company and ask them (a) where are the “moments of truth” for our customers, and (b) where are we likely to fall down on the job with them, in other words the “hot spots”? Then you can start asking “what from them?” For example, you might want to hear “This is easy for me to sign up” and you don’t want to hear “This is hard for me to use your web site”.
In my last book Your Customer Rules! Delivering the Me2B Experiences That Today’s Customers Demand2 my co-author David Jaffe and I interviewed and researched recognized global experience leaders such as Birds of Prey (Australia), Nordstrom (US), Vente-Privee (France), and Yamato Transport (Japan), and we produced a list of 7 Customer Needs that broke down into 39 Sub-Needs. Each Need and Sub-Need is written in the customer language, making it easy to mine the words and apply speech and data analytics; for each Sub-Need we also proposed its “Failure Statement”.
For example, if you and your team are concerned about achieving the 1st Customer Need “You know me, you remember me”, here are the 6 Sub-Needs and 6 Failure Statements:
Another way to catalog key words and hot topics is to take a page out of my Amazon 1999 holiday book, and list the upset expressions that you don’t want to hear from your customers such as:
- “Why …?” (a good catch since the word “why” usually connotes misunderstanding.
- “You got to be kidding me!”
- “This is the second time …” (a good indicator of experience-damaging repeat contacts, or “Snowballs”
- Contractions including “Can’t” and “Won’t”, especially when twinned with words such as “make sense” or “work for me”.
3. Capture these voices and words
There are many good speech and data engines out there from vendors large and small, so it won’t be hard for you to convince one or two of them to conduct a pilot taking your cataloged key words and hot topics from some of your customer voices and channels, analyzing their frequency and intensity (this is a place where speech analytics shines).
4. Use a recommendations engine
Now the fun begins! But first, keep in mind that speech analytics, like other forms of Big (and little) Data, it’s essential to ignore the tendency to “keep score” (like saying “Our VOC score last month was 82, a nice tick above two months ago when it was 79”) since you never know what’s driving the scores. Instead, use speech analytics as a key insight to customer experience, often a leading indicator of attrition or sales declines, and as a tool for marketing or other operational groups to improve their performance, too.
Once you have captured voices and words, you can build a recommendations engine that attempts to restore customer confidence, improve customer experience, and lead to customer re-purchase and retention. One classic recommendations engine is a coaching guide for your customer service reps, tellers, or installers “When you hear X, do Y.” More advanced, but also much more powerful, is a recommendations engine that traps near real time negative sentiment or voices and words, and reaches out to those customers before their next possible pain point with specific changes or interventions such as outbound calls from your expert technical solutions squad or a discount automatically applied to their account.
5. Learn, and compare results
This final step engages another one of the high impact aspects of Big Data, often called “machine learning”, as you build operating models. Companies like financial services companies use this learning step to see if their pricing or policy change recommendations worked, meaning that “do customers [to whom these approaches are made] buy more and stay vs. a control group?”. Or telecommunications or cable TV/Internet providers can see if their pricing or promotions worked; or utilities with their rate plans or off-peak discount programs.
If so, try it again; if not, try something else with those customers and file away that certain recommendations at certain times with certain customers did not work. This doesn’t mean that the recommendations are tossed; instead, they are stored for future use to test the predictive models.
Once you have these models in place and “learning”, it’s time to return to step 1 and deepen your speech and data analytics to improve customer experience!
1 http://customerthink.com/using-big-data-to-build-an-integrated-voice-of-the-customer-program-a-6-step-guide/, accessed 15 August 2017.
2 Your Customer Rules! Delivering the Me2B Experiences That Today’s Customers Demand (Wiley 2015). Here are the 7 Customer Needs that Lead to a Winning “Me2B”Culture; each Need breaks down into a total of 39 Sub-Needs.
- “You know me, you remember me”
- “You give me choices”
- “You make it easy for me”
- “You value me”
- “You trust me”
- “You surprise me with stuff that I can’t imagine”
- “You help me better, you help me do more”