KISS Method: “Keep it Simple with Speech”

WRITTEN BY SCOTT BAKKEN, MAINTRAX

Challenges and Priorities Survey

Speech analytics has gone from a trendy, cool technology to an essential software that reveals insights into customer-agent interactions. Speech solutions provide the tools necessary to capture, organize and analyze unstructured information to make insights actionable.

After nearly 15 years in this industry and managing more than 300 projects from a user perspective, I’ve witnessed my share of clients who have generated strong ROI by capitalizing on the data generated by pulling certain levers to improve their KPIs. Unfortunately, I have also witnessed far too many companies let their speech program collect dust because they lacked a well-prepared strategy and failed to allocate the recourses necessary. The tipping point seems to come immediately after the first application or project has been built. “Cool, but now what do we do?”

Although speech analytics can come with challenges, there’s good news as the industry on the whole continues to grow into its future state.

The number of speech vendors over the last 10 years has gone from about 20 bona fide options to more than 50, and probably closer to 100 if you include the smaller start-ups. The competitive landscape today can be broken down into five types of vendors:

  • Speech platform providers (vendors who sell via APIs)
  • Stand-alone (only sell speech analytics)
  • Speech engine providers (sell the transcription processing with no analytics)
  • Contact center infrastructure (sell WFO or the complete unified communication packages)
  • Analytics service (companies that use a speech vendor to augment their analytics business model)

If you had asked me 10 years ago to predict what the penetration rate of speech analytics would be in 2019, I would have guessed much higher than its present state around 30%. But I believe that has less to do with the maturity of the technology than with vendors being out of tune with what the customer needs. It’s really only now that vendor products and user initiatives are becoming more aligned.

The Bigger Problem: Vendors Outpacing Themselves

However, there is a big problem that doesn’t get discussed often and it’s a real elephant in the room: Many of the older legacy vendors developed analytics platforms and NLP (Natural Language Processing) tools that are so cumbersome to operate that it takes months—if not years—to build out applications and categories. At MainTrax, our experience working with more than 15 different speech tools in the last five years has allowed us to witness firsthand the complexity of building out applications such as a “first-call resolution application.” Query building can often take more than 10 hours to build a single phrase, such as “third time I’ve called.” Don’t get me wrong. I’ve made a good business out of building complex queries for clients and consuming thousands of hours developing applications, but it doesn’t have to be this way.

You also shouldn’t need a Ph.D. to operate these tools, but some products are so complex that users have to go through extensive training. Not only is this a drain on your internal resources, it leaves you completely vulnerable when your speech subject-matter expert leaves for another opportunity. I’ve seen too many companies find themselves in that pickle.

And then there are the speech vendors that hold their customers hostage by requiring them to use their professional services department—and charge them—to build speech applications despite their desire to be self-sufficient. In some cases, it seems as if vendors intentionally made their technology so complicated that you almost have no choice. Or worse, the vendor which doesn’t even give the user the ability to do so. It reminds me of having to hire an attorney to deal with the simplest of matters because the legal language was originally written to be impossible to decipher.

Avoiding Buyer’s Remorse

Speech analytics vendors tend to pile on sexy, exciting new gadgets and toss around impressive techno-speak with terms like machine learning and artificial intelligence without having a strong grasp of what their customers and prospects really need. It reminds me of a comment made by Jeff Goldblum’s character in the movie “Jurassic Park”: “Your scientists were so preoccupied with whether they could, that they didn’t stop to think if they should.” The same can be said of the developers at those companies. Perhaps the C-level people are all too willing to give their developers the green light because they may view those features as differentiators that will help to justify a stiff price bump. Too bad they seem to be in such a rush to go to market with the “next best thing” when in fact all their customers truly want from them is a more user-friendly interface and the ability to easily export their data.

It behooves the speech analytics buyer to avoid becoming a victim of a cool demo, to slow down their decision, and perform a bake-off of multiple solutions to truly understand the experience of building out an application with their own recordings. The buyer will have more realistic expectations of the tools based on their unique business requirements, and may find that a simple, high-functional tool will be more than sufficient at a fraction of the price. And we would see less buyer’s remorse.

Case in point: I recently asked a colleague why he was pivoting to a different vendor, and he replied, “We don’t need to kill a fly with a sledgehammer.”

The Corner Has Turned

Fortunately, some vendors recognize that many prospective buyers—especially small businesses—need basic blocking and tackling from products with qualities such as:

  • The ability to play in the sandbox before they buy and experience the application building, the search process and reporting capabilities.
  • A user-friendly interface and analytic platform. Can you spend one hour and understand the navigation of the tool?
  • A reasonable price. Take a look at your options and don’t assume that the more expensive vendor is better because they have more features. We at MainTrax can often perform magic while using 30% of the product features and functions.
  • Generating insights in 30 days or less. Some vendors are now saying that you can generate insights from out-of-the-box categories in one day. That’s a bit of an exaggeration, but you should be making process or performance changes based on the insights generated within the first month of use.
  • Buyer-friendly terms. You shouldn’t have to commit to a contract longer than a year.

Speech Analytics Can Help Retain Agents

In Strategic Contact’s annual report on “Contact Center Challenges and Priorities,” survey respondents often ranked agent attrition as a top challenge. Agent attrition is only getting worse with a strong economy and with the low unemployment rate. And it’s a pain felt by contact centers of all sizes.

Eric Berg, CEO of CallCenterPro Consulting, suggested that small contact centers would utilize speech to improve agent attrition (among other use cases) but the costs of the technology until now made it difficult to rationalize—even if the ROI could be established. “If there was a solution that was a fraction of the legacy vendors’ costs, small contact centers would make good use of speech to monitor new-hires during the nesting period and beyond,” he stated.

The fact that you can now buy inexpensive, powerful, yet “easier-to-operate” speech tools with impressive NLP platforms is changing the landscape, yet many are slow to adopt due to low awareness and negative historical press. With costs dropping below $50,000 (and in some cases below $20,000), the ROI certainly looks stronger if you compare it to the cost of replacing an agent at $10,000 each! This is an encouraging turn of events for small contact centers, which have been left out and want to augment their current processes with automation.

In a recent discussion I had with Wayne Ramprashad, Voci Technologies’ Chief Product Officer, he told me that Voci, which works with enterprise clients, saw the need to target smaller contact centers. He mentioned that they “…can now help smaller contact centers generate insights quickly with powerful tools at a reasonable price. Value for the cost is key. Our analytics platform is powerful, and we’ve had clients who pivoted from legacy providers to our solution with delight by easily gaining insight into the voice of the customer. The Enterprise base might not be completely saturated, but it’s the SMB market that is taking action and amounts to a high number of seats.”

Key Trends

Here are a few other trends that I see in the market:

  • SMB contact centers are also looking beyond agent performance and agent retention by focusing on customer experience (CX). Fortunately, speech tools are a natural fit to improve Csat.
  • Speaker separation (channel separated recordings) is enabling speech tools to gain deeper insights into agent-customer interactions. Some vendors can also separate agent and customer channels on mono recordings using a feature called diarization.
  • In the last two years alone, more than 30 new speech vendors have entered the market—many with innovative technologies. The industry has become a bit of a bloodbath.
  • Omnichannel and call journey analytics are gaining momentum. I view this as welcome news because all interactions and touchpoints need to be considered to truly understand the customer.
  • The lack of vendor professional services continues to hinder buyer satisfaction and speed to insights.
  • Real-time speech analytics solutions, which analyze phone calls while they’re in progress, can alert a supervisor to intervene and resolve important issues before they escalate—or prompt an agent to try a different approach, switch to a different script or attempt a specific upsell. However, real-time analytics is difficult to utilize for automation. Furthermore, this assumes that the technology is working perfectly. In my view, this technology needs to mature before it performs as well in practice as it does in theory. Yet even if the technology performed as advertised, questions remain. For instance, do you have the internal resources to capitalize on call alerts in progress?
  • I expect we’ll see a separation of vendors: Those which offer less expensive, simple, yet powerful technology that enterprises may label as “not good enough,” and vendors investing in predictive analytics, machine learning, and advances in AI in order to sell to the buyers that want cutting-edge analytics. With that said, both types of vendors will continue to gain momentum. Uninformed buyers who do not understand the differences in technologies will struggle in the procurement process and unfortunately and will be prone to selecting whichever vendor presents the best demo or proof-of-concept readout.

Remember to KISS and Forgo the Expensive Add-ons

Legacy vendors have created some raving fans and some amazing case studies. But in the vast majority of cases, their customers didn’t need all the fancy, expensive bells-and-whistles to accomplish their goals. In fact, those extravagant features probably got in the way. Insights could have been generated quicker, cheaper and with fewer resources.


9 Leading Practices for Selecting a Speech Analytics Tool and Vendor

If you’re considering investing in speech analytics, these nine tips will help you find the speech tool that best matches your needs and maximizes the business value of your results:

  1. Don’t get caught up in prolonged discussions about accuracy. Initial high accuracy rates can be a bonus, but for some companies, accuracy isn’t even a top-three criteria for choosing a particular speech software. And as Roger Lee of Gridspace recently said, “Stop concerning yourself with statistical significance.”
  2. Don’t go overboard with feature requests in your RFPs. Just because one vendor offers something doesn’t mean you need to ask another if they can boast the same claim. Is HIPAA compliance really needed? Will your organization actually benefit from an Arabic language pack?
  3. Be practical. A speech tool’s standard features and functions should be clean and intuitive and satisfy the bulk of any user’s functionality needs. Consider how you’ll solve specific problems around targeted business issues. With the right tool and easy-to-interpret reports, beginning users should be able to generate basic but powerful insights within the first week of deployment.
  4. Consider your 12-month goals and how the software will help you meet those goals. Ask about the availability of standard and custom reports along with workflows that will support your progress towards these goals.
  5. Determine whether you can integrate metadata with your speech data. Do what you can to maximize the value and usability of speech insights by incorporating metadata, which is data that provides information about other data. For example, intelligent call-recording systems have the ability to tag valuable data such as account identification, product name, and dollar value of the sale from virtually any servicing system (e.g., CRM, ERP, helpdesk, homegrown). You can then use all that additional data in your call-recording searches in your reporting to create high-value business intelligence.
  6. Identify what support and training are included. Ask each of the various speech vendors you’re considering for a detailed overview of their post-sale support. That should include a dedicated end-user representative, a technical support person, support hours from their help desk and documentation about technical and user training. Make sure they differentiate between training for business users, technical users and system support.
  7. Don’t fall prey to an impressive demo. Get clear about what high-value insights you desire for your business and ask to test the system against a subset of your own calls. Consider conducting a “bake-off” with multiple speech vendors using your own recordings. This can take additional time, but you’ll gain valuable knowledge about exactly what you’re purchasing.
  8. Don’t jump onto the name brand bandwagon until you’ve done an unbiased, thorough evaluation to ensure that they are good fit. You’re not “playing it safe” as much as you may think by choosing a new product from a nationally known Fortune 50 behemoth.
  9. Ask to speak with references. And call them.

About MainTrax
MainTrax is a visionary team of experts uniquely qualified to help companies maximize speech analytics technology and transform customer data into smart, actionable business solutions.


Scott Bakken, Founder and President of MainTrax, is a highly respected independent voice in the speech analytics industry. His company provides professional services that help end users use their existing speech analytics tools to deliver actionable business intelligence. Bakken was named an Ernst & Young Entrepreneur of the Year finalist.