Al Kingsley MBE is CEO of NetSupport, Chair of a multi-academy trust in the U.K., tech writer, speaker & author of multiple education books.
The rush toward AI adoption is still strong. Companies worldwide are racing to integrate it into their offerings or launch new AI-based solutions. However, not every new solution will succeed—far from it—as MIT reports that 95% of the 30,000 AI products introduced each year will fail.
The words "now with AI" or "AI-based" may be enough to spark prospective customer interest, but you need more substance to stand out. If the product doesn’t deliver something new, useful or productive for users, then there are plenty more solutions in the sea.
So, how can you ensure your product is among the relatively few that ride the waves and avoid a costly misstep for your business?
Do your research.
Consider this. A company launched an AI-based "floristry helper" app for people to ask it about the flowers they see in the shop before they request a bouquet. The company envisaged that people would ask the AI which flowers to choose for any occasion. It would then make recommendations and answer questions about seasonality or durability.
But here’s the thing: It just didn’t take off. Why? Because floristry is a people-first sector. The friendly human florists in the shop have a wealth of knowledge and—critically—experience, which people were already accessing with face-to-face conversations.
The app failed because the company had not done the necessary research to validate what it thought was a solid idea and there was actually no problem to solve at this time.
The lesson here is that you need more than a gut feeling to start producing a product. Just making it "AI-based" won’t automatically mean success if the premise isn’t on point.
Doing the necessary research to define the real user problems that you could answer with your AI product will ensure that your development starts off on the right foot. Surveys, pilot studies and talking to your prospective audience can all indicate whether your idea will fly.
Establish trust first.
If your product is a new kid on the block, then you need to work extra hard to establish your credibility upfront.
Without the prior achievements or successful products to demonstrate your capabilities, prospective customers will often need more proof or persuasion before spending their hard-won dollars on trying your product.
These days, data security and privacy are among the principal concerns for any organization operating online. In terms of AI, you’ll need to consider how you’ll address this and establish clear data governance practices.
You’ll also need to be clear on how you deal with AI’s bias. The critical part is to be transparent about all of it to give prospective buyers a chance to see that you are serious about your offer.
A great way to provide a foundation for trust is to align your company with industry-level security standards such as SOC 2 and/or ISO 27001. So, look carefully at these and others. These standards are well-respected worldwide, and the trust you earn from them is often well worth it, even if achieving them takes considerable time and effort.
Future-proof your product.
With the breakneck speed of transformation, every AI product needs constant attention and updates. But your users should not see that complexity. Your interface should be clean, logical and human-friendly. Think about how uncluttered chatbot start screens are despite the mass of complexity going on in the background.
In addition, the tech behind your user interface (UI) should be reliable and responsive—so comprehensive testing is key. As I’ve said, in a world full of other AI solutions, customers won’t hesitate to go elsewhere if your product causes them the slightest hiccup.
Don’t forget about the non-tech side of things, too. Work with your users, listen to their feedback and nurture relationships. After all, you’re making this product to help them achieve their goals. Relationship-building is so important if you want people to choose you and, equally importantly, to stick with you.
On that note, consider whether you can integrate complementary products from other companies. Acknowledging that others may "do" part of what you do differently or that a different product is used as an industry standard doesn’t necessarily mean you are locked out of that market. Giving consumers the flexibility to best use your product within their workflow demonstrates your understanding of them and their sector.
For example, at NetSupport, we integrate the safeguarding module of one of our education solutions with two other mainstream safeguarding recording products, both of which are used widely in schools. It doesn’t mean people don’t buy or use our product. Quite the opposite: Customers appreciate the value of flexibility and of having the strengths of both solutions in their toolkit.
Sell, support and scale.
Even the best product will fail if you can’t sell, support or scale it.
Honest marketing (without inflated claims), responsive technical support (bot-free usually wins the day) and the ability to roll it out across an organization’s network without bringing everything to a standstill are the very least you can do to move your offer forward.
Your success hinges on being able to deliver what you promise. Even if your AI product is currently exceptionally simple, make sure it does its thing exceptionally well and market it on the strength of its simplicity.
Also, ensuring your sales and marketing messages align will mean prospective customers get a unified picture of what it can do for them—and having a knowledgeable team behind the scenes (e.g., developers, educators, security specialists) to answer any questions will all help you to build trust.
Keep it real.
Ask yourself these basic questions: Are we solving real problems? Do we have trust built in? Is our tech flexible? Can we deliver what we promise?
If you can answer them honestly (or are working on it), then you have a solid foundation for your product’s future success.