Co-founder and President DigitalGenius
Share this content
Robot face
Hub Icon

AI in the contact centre: How to research your options and reach out to vendors

17th Jan 2019
Co-founder and President DigitalGenius
Share this content

As a customer service professional, your goal is to help your customers by solving their concerns quickly and effectively. This isn't easy to accomplish if you're spending half of your day on repetitive processes that slow down progress. Recent advances in technology make it possible--here and now--to manage those tasks.

In his book, AI is my Friend – A Practical Guide for Contact Centers, Mikhail Naumov, co-founder and CSO of DigitalGenius, reflects on his years of experience to show readers how to transform a contact centre with practical artificial intelligence.

In the following extract, Mikhail shares the steps and precautions a company should take when identifying potential AI solution vendors to engage with for their support centre. 

Researching your options

The first step in implementing AI is to investigate the available options. There are many companies making a lot of noise, including all the early-stage dot-AI startups, well-funded research organisations, incumbent software vendors, and the Narrow AI Specialists.

Depending on your appetite and risk profile, you may want to research each option. This discovery step usually begins with basic internet searches. Try key terms such as “AI for contact centres,” “call centre AI,” or “contact centre automation.” This will help you gather as much diverse and relevant information as possible to narrow down your set of potential vendor candidates.

Also, utilise reports from the analyst community, such as Gartner and Forrester. These reports have traditionally covered mature technologies, but they are beginning to cover AI as the fundamental value of AI becomes clearer.

The analyst community anticipates that AI could become as significant as the cloud or the internet itself, so they are responding quickly through increased research and thorough analysis.

Most companies already subscribe to these research databases, so the reports tend to be highly accessible. Annual reports highlight core technologies, so look for recent editions and market guides that focus on AI specifically for contact centres.

Reaching out to vendors

The second step is to make a short list of companies you plan to reach out to directly. Contacting every single AI company on the market is unreasonable and time consuming; there are simply too many of them now. To narrow down the list, create a set of relevant criteria and make a list of companies that meet your criteria. Also, think about creating a diverse list, so you might have some dot-AI startups doing cutting-edge work that you want to know about, along with some bigger companies and narrow AI specialists offering proven solutions.

Keep in mind, the information you gather on AI vendors online might already be outdated. This market changes rapidly. If you come across a white paper or case study about a specific product, it’s likely the product has evolved since publication. AI companies pivot and evolve often, so what you see in a white paper might be a historical case study and not the most current representation of products the company offers. Careful diligence during online research matters, because it will help you eliminate some companies from your list. It will also help inform your conversation when you speak with the vendors.

Fortunately, companies are easy to contact. Websites will almost always allow you to submit a web form or schedule a demo. These AI companies are highly motivated, so they don’t want to miss the connection with a potential customer. Once you make contact, the company will help to validate your needs and see if their solutions are a good fit.

Only a company that has fine-tuned their understanding of their ideal customers will begin asking questions even before presenting you with their product.

Bear in mind, an established company that knows their market will turn the tables on your first phone call very quickly. They will start asking you about your contact centre. How many agents do you have? What’s the volume of incoming messages? How many languages do you support? What time zones are you in?

Don’t be alarmed! This can feel strange since you contacted them to get more information about their product, but it’s your first clue as to the type of company you are dealing with. An amateur startup most likely won’t do this.

Only a company that has fine-tuned their understanding of their ideal customers will begin asking questions even before presenting you with their product. That’s because they want to make sure your company fits their checklist.

They also use thoughtful questions to prove to you that they truly understand your pain points and the requirements of your business. When this happens, understand that you are dealing with an experienced vendor.

This is quite different from what you’ll hear on a phone call with an early-stage dot-AI startup that probably doesn’t yet know which questions to ask, and is just happy to have you on the phone in the first place. If you submit a web form and receive an immediate response, but the vendor doesn’t ask any questions, that’s probably an indication that the company is so early in their life cycle that they don’t know who their ideal customer is yet. In an ultra early stage, companies take anyone they can get. Also, if they don’t ask thoughtful questions about your customer service operation, that might indicate that they have a limited understanding of your industry.

And finally, in terms of the incumbents, you probably won’t have to think about contacting them. They’ll call you first and offer an AI expansion to the existing software you’re already licensing from them.

Once you’ve researched, created your short list, contacted the companies, and noted their responses, the next step is to schedule demos. You need to see the products in action. In a demo call, someone from the company will demonstrate their product via an online screen-share, or they might set up an in-person meeting. During the demo, they will attempt to show that they understand your business and prove to you that their product solves your specific challenges.

At this point, the company should tie the product back to key performance indicators (KPIs) or business objectives that your contact centre is trying to achieve. Look for this information at some point during the demo. If the person on the other end of the call is just pitching you the product, if they fail to tie it back to your KPIs, chances are they haven’t really considered how their product directly relates to your success.

They might be very excited about it, and it might sound like the greatest product on earth, but be wary of continuing this kind of relationship. There’s a high likelihood that their focus will remain on their product, rather than on how it serves your specific business.

You need a company that understands your industry, a company that is straightforward and truthful about whether they offer a product that can help you. In the interviews and demos, work your way through your short list. Eventually, a few companies will stand out.

After evaluating your short list to identify the stand-outs, create a memo or presentation for your team or budget holders.

AI is my Friend – A Practical Guide for Contact Centers, available on Amazon, describes how customer service professionals can leverage Machine Learning software to enhance agent performance by selectively automating the repetitive parts of the workflow by classifying, prioritising, and automating responses service inquiries, emails, and messages, freeing up agents' time to focus on personalisation and more, high value-added service. The book features real life stories from customer service professionals who adopted an AI solution, and reflects industry learnings and actionable insights for those that are ready to make the first step towards adopting AI.


Replies (0)

Please login or register to join the discussion.

There are currently no replies, be the first to post a reply.