The Secret to Successfully Demoing Your AI Features

July 2, 2024
Stop asking skeptical buyers to imagine — show them what your AI can do.
Meghan Spork
Table of Contents

It already feels like AI is everywhere — and yet we’re only at the tip of the iceberg. 

There’s no denying that that number of AI tools available is going to keep growing and that this technology is reshaping the way we do business. No matter where you sit in the organization — you’re either developing your AI strategy or signing up to get left behind. 

And yet we’re seeing software companies developing these amazing new AI features and then running head first into a critical challenge — they can’t actually show these new capabilities to buyers in the sales process. 

To illustrate why this is important, let me tell you a story … 

A Tale of AI and the Skeptical B2B Buyer

It’s late May and Salesforce is reporting their Q1 earnings. 

Marc Benioff shares that Salesforce is all in on AI — “It’s the beginning of a massive opportunity for our customers to connect with their customers in a whole new way with AI.” 

But he also acknowledges that many customers are still struggling to understand AI’s potential and practical applications. And that this gap in understanding presents a big hurdle for Salesforce and any other company that is trying to showcase their advanced AI features. 

Benioff emphasized that explaining AI and showcasing its potential impact requires a different skill set than what it takes to sell marketing, sales or service software. 

This resonated with me because it’s something we hear every day. Companies are building these incredible Generative AI features and can’t showcase them in a live sales demo to enable buyers to really understand the value. 

A prime example is the team at Gong. They’ve created an incredibly powerful AI functionality — and yet weren’t able to show it to their buyers in live demos and have confidence that the AI would trigger the right responses. 

B2B buyers are skeptical. They need to see it to believe it. By visually demonstrating AI in action and its tangible benefits, we can help buyers cross the chasm and really grasp the impact AI can have on their business.

And with tighter budgets and new purchases facing more scrutiny than ever, GTM teams need to be able to showcase their most powerful, value-driving features to instill confidence and get deals done. 

Why AI Is So Hard to Demo And How to Fix It

So, what makes showing AI features, specifically GenAI, this difficult? Great question. 

It’s the sheer amount of data required for AI models to produce reasonable results. 

Take it from Alan Chan, Senior Manager of Sales Engineering at Gong, “Because our product requires a ton of data to trigger the right AI responses, it has been very challenging to have a demo environment that is reliable and keeps up to date with all our feature releases.” 

Depending on the data type, data creation for demo instances is already a challenge for many solutions engineering teams. And now AI has upped the ante, requiring data creation at scale and often requiring data to flow into the platform on an ongoing basis. 

To combat this, we see organizations take a handful of approaches to solve the issue with varying levels of success:

Ask buyers to imagine. 

For organizations who are just starting down this journey, we often see them rely on doctored screenshots and product gifs to paint the picture. They often still show the feature in the product, and ask buyers to envision what the tool can do. “Imagine you ask the AI a question, like X, and it would spit out a result, like Y.” 

A key downside to this approach is that it introduces doubt and risk into the buying cycle. If your AI is one of your key drivers of differentiated value, not being able to show to buyers in live demonstrations is not ideal.

Show production environments. 

Sellers are gonna do what sellers are gonna do … especially when there’s a deal on the line. We’ve talked to a handful of companies who’ve admitted that their sellers or solutions engineers would show production environments or the organization’s own environment to buyers who were especially interested in seeing the AI features in action.

But again, this approach opens up quite a bit of risk. Showing prospective buyers production environments or even your own environment can open up issues with SOC 2 and other privacy regulations. And there’s no worse way to instill trust in your privacy and security posture with potential customers than being lax with the data of your current customers.

Anonymize existing customers’ data.

A workable option that we’ve seen only a few organizations successfully employ is  anonymizing data from some of their top customers. Specifically, they’ll take a customer instance and put their data through a filter to strip out or mask any key identifying info. While it’s certainly a better option than showing production environments, it’s one that takes a lot of time and still carries some risk.

Solutions engineers still have to do the work on anonymizing the data and adding it to the demo instance — and while it’s possible, this work often isn’t scalable. As a result, it typically happens at a certain point in time and then that data goes stale. And it’s noting that anonymizing customer data still introduces a certain level of risk. If enough of the right attributes are left intact, there’s a risk someone could ID the customer — which doesn’t bode well for either that customer relationship or the future one you’re trying to build. And anonymizing data can reduce the richness and thus usefulness for some AI models.

Use synthetic PII-free data.

Last, but not least, organizations look for solutions - like TestBox - to create high quality, synthetic data models that enable their demo instances to show off ALL of their value-driving product features, but especially their AI.

We work with customers to understand their ideal data story, inject that data into their live product and then develop an activity plan to create that data on an ongoing basis, ensuring it never goes stale.

While it does take time (roughly six to eight weeks) to stand up the solution, the ongoing benefits are massive. First and foremost, it empowers your customer-facing teams with the ability to show buyers and customers your AI working as expected (as well as other value driving features). This approach also reduces the lift on your solutions engineers, who no longer have to find ways to generate the massive amount of data to power the AI or continuously anonymize customer data. And finally, there’s no risk. Because the data is synthetically generated, it’s as good as customer data, but without the risk of any privacy or security concerns. 

Stop Asking Buyers to Imagine — Show Them What’s Possible

The whole point of a demo is to prove to your customers that your product really does what you say it can do. Even though AI is quickly becoming an important part of almost every product’s feature set, most buyers are still skeptical and need to see it in action to believe it. Ensure your revenue teams have the right tools to show off your AI to its fullest capabilities.

When you accurately show the power of your platform, you foster trust with buyers, which, in turn, creates a positive downstream impact on your revenue results: higher win rates, shorter sales cycles, and, most importantly, happier customers.

If you’re struggling to properly showcase your AI features, reach out to our team today to find out what we can do to help.

You might also like

No items found.

Watch our Chief Solutions Officer, James Kaikis, talk about the future of solutions based organizations

Watch the presentation to learn about the change that you can make within your organization.

TestBox Logo

More revenue, less effort

Spin up live demo environments, trials, and POCs in seconds with TestBox. Close more deals and speed up your sales cycle — all with fewer resources than ever before.