There is a big problem when discussing artificial intelligence (AI) with a prospect or customer. With AI, and positioning AI, being a relatively new market with a broad coverage, navigating AI positioning without knowing a person’s level of knowledge can be tricky. To make matters worse, MSP’s own sales and marketing teams may lack sufficient knowledge of AI.
If we just start at the very basic level, is there a difference between machine learning, deep learning (DL), and AI? If so, what are the differences – and do they matter?
Let’s then go a bit higher up – what are the differences between AI, generative AI (GenAI), agentic AI and AI agents? Again, does it really matter?
In many ways, yes, it does matter – the difference across all of these different approaches to dealing with data and processes can make or break how a user sees the success of any AI usage. Also remember that AI is a highly dynamic market: just a year ago, few had heard of terms such as GenAI, agentic AI or AI agents. In a year, experts might start using a whole new set of AI terms. Also, bear in mind that there has been a lot of ‘AI-washing’ out in the market: vendors have taken existing data analytics products and rebranded them as ‘AI’. They are not, and buyers are getting wise to being mis-sold solutions that cannot adapt to how the business needs to operate.
However, at a better level, the differences do not matter at all. Why? This is because AI in itself is unimportant – it is the end result that counts.
Don’t lead with AI
Sending in a salesperson who engages with a customer along the lines of “We have deeply embedded AI that will solve all your problems!” is just asking for the prospect/customer to call the salesperson’s bluff and demand distinct and discrete proof of what they are saying – proof that the salesperson may well not have. Should the salesperson then spout on about how the embedded AI uses the GenAI that has been trained on large language models (LLMs) to create agentic AI entities that can replace the helpdesk staff, that salesperson should expect to be shown the door. Equally, teams should quickly discard any marketing material that falls into the same trap, as it works against actual sales.
There is also the problem where a salesperson, driven by targets, thinks that they can use AI arguments to baffle the prospect into a purchase – but then finds themselves up against someone who does know AI to a greater extent. All of a sudden, the salesperson is then faced with comments such as “What LLMs is this AI trained against – am I protected against intellectual property and copyright theft when using it?”, or “This seems to me to be far more of an AI agent than agentic AI – can you talk me through it in detail?”. The sale is then all too easily lost.
Start with the pain point
Instead, if the salesperson asks, “What problems are most pressing to your business right now?”, the conversation becomes far more meaningful and productive. It is outcomes that define how successful a business is going to be – not the technology.
The salesperson needs to be listening for things like “Our customers are unhappy with our issue resolution rates/times”, or “We are struggling to identify the correct prospects to bring onboard as customers”. These are business problems where AI can certainly help – but, as with all the other technologies that have gone before, AI of itself is no silver bullet.
Sure, AI can speed everything up – and if the underlying processes are bad, it will just result in faster, more harmful outcomes.
How MSPs can talk AI the right way
To avoid the myriad minefields of AI, what should an MSP be doing with its sales and marketing divisions?
- Focus more on messaging around outcomes, avoiding too much focus on the AI aspects of the solution.
- Marketing should develop and present AI use cases to prospects and customers that address common business needs like prospect attraction, customer support, and issue resolution.
- Educate the sales and marketing employees on the basics of the different types of AI – not to any great depth, but enough to be able to engage with customers who want some detail.
- Ensure that salespeople refer back to more AI-savvy technical staff in the office where discussions become more deeply AI-specific.
- Do not overpromise. Under achieving on statements made around how AI would help a customer can be catastrophic. AI is not a cure-all, and its wrong usage can quite easily cause an organization to lose customer confidence and collapse its business.
- Do not over-depend on AI internally. Many organizations are adopting AI to help in their approaches to the market. This approach works, as long as teams carry it out correctly—with strong human oversight to ensure they achieve the desired outcomes. A welcome by-product of this is that you are then learning in-house what you are trying to sell to the customer: as the saying goes, you are ‘eating your own dog food’.
Selling outcomes, not just AI
Finally, this highlights the need for a more consultative approach to sales. AI is not a standalone solution; rather, it is a tool to achieve outcomes more quickly and, hopefully, with greater accuracy. To leverage AI effectively, customers may need to change their existing processes, potentially quite drastically. By helping customers improve their processes before implementing AI, you can achieve better sales outcomes and unlock new revenue streams.
In summary, while AI is not the complete answer, it is an important part of the solution. The most crucial aspect is ensuring that the customer receives what they truly need.
Photo: Golden Dayz / Shutterstock
