What AI Actually Changes in Sales 

Insights by Lasse Teeriaho, Co-Founder & Pre-Sales Architect, Forward Forever 
 

I have been part of Forward Forever since the early days, first as one of the founders and as managing director, then gradually shifting my focus towards clients and technical work. Recently we decided to change my angle again: I now concentrate on sales, working as a pre-sales architect and building AI-assisted tools for our own sales and marketing. 

It is an unusual combination, but that is exactly why it works. I build the tools, and I use them myself. That feedback loop tells me very quickly what is actually useful and what just sounds good in theory. 

Start by freeing up time, not by chasing the perfect tool 

The most obvious place to start with AI in sales is automation of routine tasks. Meeting memos, proposal drafts, follow-up summaries: these are things AI handles well, and for most salespeople they represent a surprising amount of time every week. 

Teams transcripts alone open up a lot. After a client meeting, you can generate a structured memo, pull out action points, or start a proposal draft, all from the transcript. Things that used to be postponed until Friday afternoon can now be done in minutes. 

But here is the part that often gets missed. If you free up time without deciding what to do with it, nothing changes. Building the tools is only half the job. Sales leadership needs to think in parallel about where that recovered time should go, because it will not direct itself. 

Build for the whole organization, not just for yourself 

One of the clearest lessons from building these tools internally is how easy it is to build something that only works for you. AI-assisted tools come together quickly, and that speed is also a risk. You solve your own problem, the tool fits your own way of working, and then it turns out nobody else wants to use it. 

The moment you start building, you need to keep the whole organization in mind. What does the workflow look like for someone else? Where does it break down? A tool that one person uses enthusiastically is not the same as a tool that scales. 

This is also why showing clients our own setup works well in practice. We are not presenting a theoretical framework; we are showing them something we actually use. That makes the conversation about their needs much more concrete, because the use cases in sales and marketing are broadly similar across organizations. We do not start from scratch each time. 

Lead generation is more useful when the reasoning is visible 

AI-assisted lead generation is one of the more valuable applications in sales, but it only works if the output can be trusted. When a lead surfaces automatically, the salesperson looking at it needs to understand why. What signal triggered it? A job change, a previous contact, a specific company characteristic? 

Without that reasoning being visible, you get a list that nobody acts on. With it, you get something a salesperson can actually evaluate and prioritise. The difference between those two outcomes is not the AI; it is how the tool is designed. 

The CRM is moving to the background 

One shift that does not get enough attention is what happens to the CRM interface as agents and automations take over more of the routine work. If an agent is logging call notes, updating pipeline status and triggering follow-up tasks automatically, the salesperson no longer needs to live inside the CRM the way they used to. 

Salesforce made a significant move in this direction with their Headless 360 launch earlier this year. The same philosophy applies to the Microsoft stack. There is nothing stopping organisations from building CRM logic that runs in the background, surfaces only what the salesperson actually needs, and lets agents handle the rest. The interface becomes less important when the process runs behind it. 

Marketing gets a lot from what already exists 

One pattern we keep coming back to in marketing is the gap between expertise and content. Most expert organisations have the knowledge; they just do not have a reliable way to turn it into published material without it becoming a burden on the people who actually know things. 

We do quarterly reviews with clients going through what is new in the Microsoft ecosystem. That session, which happens anyway, contains enough material for blog posts, LinkedIn updates and other content. With the right setup, the recording does the heavy lifting and the content team can take it from there. The experts do not need to sit down separately to write anything. 

That kind of flow, where the content comes out of work that is already happening rather than being added on top of it, is where AI assistance makes the most practical difference in marketing. 

What actually motivates me about this work 

What I find most interesting about this role is how quickly you can see whether something works. I am not just building tools in isolation; I am using them in real sales situations. If a tool is genuinely useful, I reach for it after the next meeting. If it is not, that becomes obvious fast. 

That kind of immediate feedback is rare in most development work. It keeps the work grounded, and it means we end up building things that actually get used rather than things that looked good on a whiteboard. 

Lasse Teeriaho is a Co-Founder & Pre-Sales Architect at Forward Forever, where he builds AI-assisted tools for sales and marketing and works with clients on Microsoft-based solutions. 

Thinking about AI in your sales process?