I opened my feed this week and saw a buyer intelligence acquisition, a new AI agent for tracking brand visibility, and a stat that agentic AI spend has crossed $200 billion. That’s a normal week now. It’s like standing in the middle of a car dealership lot where every single salesperson is running at you at once, each one convinced their car is the one you need. You haven’t even said what you’re shopping for yet.
So within the last year, as so many new AI tools have popped up trying to make a name for themselves, I stopped reacting to what a tool does. I ask what it’s actually for, in my world, with my customers, at the moment they’re actually in. If I can’t answer that, I don’t care how good the demo is.
Let me walk through how I actually run something through that filter, using this week as the example.
Start With the Problem, Not the Product
Here’s what usually happens. A tool lands in my inbox with a slick capability. AI-powered this, agentic that. And my brain wants to get excited about the capability before I’ve asked whether I even have the problem it solves.
It’s the same mistake as showing up to play pickleball without knowing a single rule, and getting matched up against a 65-year-old woman in a visor and compression socks who lobs the ball right over your head while you’re standing in the kitchen wondering what “the kitchen” even means. She’s not even trying to embarrass you. She thinks it’s a teachable moment. There’s zero chance you’re winning that game as a newbie who skipped the rules, and there’s even less of a chance of it if you also showed up with an outdoor ball because nobody told you indoor and outdoor pickleballs aren’t the same thing either. Wrong ball, wrong game, and you’re getting run off the court either way.
These pickleball people are serious. It’s time we got just as serious about what tools we’re actually using and how we plan to use them before we ever step on the court.
So I make myself stop and write the problem down first, in my own words, before I look at the tool again. Not “we need better personalization.” Something I could say out loud in a sales meeting, like: our reps don’t find out an account is in-market until they’re three calls deep into a cold sequence. Or: I have no idea if our content is even showing up when someone asks ChatGPT a question we should be answering.
If I catch myself using the vendor’s language to describe my own problem, that’s the tell. I don’t have a problem yet. I have a pitch stuck in my head, and I need to shake it loose before I go any further. Otherwise, I’m just the newbie standing in the kitchen, holding the wrong ball, wondering how I lost so fast.
Here’s the simple format I actually use when I’m evaluating a new solution.
Have You Actually Defined Your ICP?
This is the step almost everyone skips, and it’s the one that makes every decision after it so much easier.
Take what just happened with Zoom. They announced they’re acquiring Common Room, a go-to-market intelligence platform already running inside companies like Atlassian, Autodesk, and Snowflake, built to give revenue teams a real, person-level read on a buyer before a rep ever dials the phone. My first reaction was, That’s genuinely smart. My second reaction was, it’s only smart for someone who already knows exactly who they’re chasing.
I’ve seen this go wrong. Handing a tool like that to my sales team with a fuzzy ICP is like using a GPS but plugging in an address that just sounds close enough to where you’re actually trying to go. It’ll still talk with total confidence. Recalculating. Turn left in 500 feet. And it’ll drive you right past every customer who’d actually buy from you, delivering you somewhere else entirely with total certainty the whole way. That’s not insight. That’s just being wrong faster, with better graphics. Define the ICP first. Then let the tool go find you more of exactly that. Not before.
What Objections Are You Actually Hearing?
Every vendor pitch promises to remove friction. So I ask myself, whose friction, and did I actually hear it, or am I imagining it because it sounds like a real problem?
If the sales team’s real objection is “I don’t trust this lead score,” a fancier AI scoring model doesn’t touch that. I think about this the same way I think about teaching someone to read. Handing a struggling reader a thicker book with a nicer cover doesn’t make them read faster. It just puts more distance between where they are and where they need to be, and now they feel worse about it, too. You have to go back to what’s actually breaking down first.
But if the objection I keep hearing is “I don’t know if our content still works, I feel like it’s disappearing,” that’s telling me something real, and it’s worth chasing down. That second one is actually happening right now. AI Overviews are already cutting traffic to top-ranking pages by more than a third in some cases. That’s a page-one SEO strategy quietly losing its return, the way a store loses foot traffic once a new highway bypasses Main Street. So when a tool like Profound’s new Aim shows up, built specifically to watch how your brand gets cited inside ChatGPT, Gemini, and Perplexity and turn that into next actions , I don’t dismiss it as noise. I check it against the objection I’ve actually heard. If nobody on my team has raised that concern yet, I let it wait. If they have, I move.
What Questions Are You Actually Asking?
Before I bring anything to leadership, I make myself sit with three questions and answer them honestly, out loud if I have to.
- What decision does this actually help me make faster or better? (Not what task it automates, but what decision it improves.)
- What happens if I skip it? (If my honest answer is not much, that tells me everything.)
- Who on my team is going to open this thing every single day, and do they even want it?
This is the one leaders get wrong the most. I’ve watched brilliant, idea-rich leaders throw four or five new tools and initiatives at a team in a single quarter, all genuinely good ideas, and the team ends up more confused and less productive than if they’d gotten one clear priority. Big ideas with no plan to execute don’t move a team forward. They just take everyone’s eyes off the one thing that actually mattered.
If I can’t answer all three cleanly, I don’t kill the idea; I just table it. I come back once the answers are clear, instead of forcing it.
Where Are Your Customers in the Buyer’s Journey?
This is the question that cuts the most noise out of my day, because most tools are built to solve a single moment and marketed as if they solve every moment.
I think about this the way I think about wearing stilettos on a hike. Both are shoes, sure, but one of them is going to leave you with a twisted ankle 100 feet into the trail because it was built for a completely different terrain. A buyer intelligence tool is built for the early stretch of the trail, before a rep has even said hello. A conversational AI tool is built for the middle of the hike, when someone’s actively comparing you to their other options and wants a real answer fast. An AI search visibility tool works even earlier than both, shaping whether you’re even on the map before anyone’s put you side by side with anyone else.
Agentic AI is starting to blur all three of those moments together, and the numbers back that up. Gartner expects 40% of enterprise applications to have AI agents embedded by the end of this year, up from under 5% last year. Read about that here. That’s a real shift, and I’m not dismissing it. But I remind myself, agentic isn’t a stage of the journey; it’s how the tool moves, not where you’re going. I ask where the actual gap is in my funnel first, then I look for the tool built for that stage. Not the one with the flashiest agent demo.
The Gap Nobody’s Budgeting For
Here’s what most of this week’s coverage isn’t saying loud enough. AI agents are increasingly making targeting, suppression, and segmentation calls, and many of them are running on data nobody has opened in years. Consent records, lead-scoring rules, and suppression lists that are technically still active but haven’t reflected reality for a long time. And right now, only 19% of organizations are even tracking KPIs on their generative AI use. Check out the article.
I love a beautiful watch, and here’s the thing about them. A watch that’s running five minutes slow is worse than no watch at all, because you trust it completely and still show up late, fully confident you’re on time. That’s what bad data does inside an AI tool. It doesn’t announce that it’s wrong. It just keeps telling you the wrong time with total confidence.
I say this to every team I work with. Buying tools without correct data will just make you wrong faster.
So before I let any AI tool touch my revenue motion, I go audit the data it’s about to run on first. That part doesn’t make for a good headline. But it’s the part that decides whether everything else actually works.
The Filter, Simplified
When a new AI headline crosses my feed now, here’s what I actually ask before I let it take up any more of my time.
- What problem does this solve, in language I could say out loud?
- Does it match a customer I’ve clearly defined?
- Does it answer an objection I’ve actually heard from my team or my buyers?
- Can I name the decision it improves, and who’s going to use it?
- Do I know exactly where in the buyer’s journey it belongs?
If it clears all five, I give it a real evaluation. If it clears one or two, I bookmark it and move on. And if it clears none of them, I let it go, no matter how many headlines it’s generating this week.
Here’s the part worth sitting with. Your buyers are running this exact same filter on you right now, whether they’ve ever put words to it or not. They want to know you understand their problem, that you actually get who they are, that you can speak to the objection they walked in with, that you’re solving for the decision that matters to them, and that you show up right where they are in their own buying journey. Get clear on your own filter, and you make it a lot easier for them to choose you when they’re running theirs.
Need help developing an ICP or understanding your buyer’s journey before you bring another AI tool into the mix? That’s exactly the kind of thing I help teams work through. Reach out to robo@robinbowling.com and let’s figure out what’s signal and what’s just noise for you.
Living Positive on Purpose 