I have spent twenty-five years selling data, and in all that time I have never once believed the word "better." I have said it. It is in every deck, every battlecard, every talk track ever handed to a new rep. Ours is faster, cleaner, deeper, more accurate, more complete. Better. I said it because the room expected to hear it, the way you say "fine" when someone asks how you are. But I never believed it, and it took me a long time to work out why.
Here is the reason, and it is not a matter of taste. "Better" is a claim about a single number. It says: line these things up on one axis, and mine sits higher. That claim is coherent for a very small class of things. A longer tape measure is better than a shorter one if what you need is length. A brighter flashlight is better in the dark. When the thing you are buying has one job and one dimension, "better" is a fair word, because there is genuinely only one direction to move.
Almost nothing I have ever sold is like that. And almost nothing worth an enterprise signature is like that. What I sell is a match between a dataset with dozens of properties and a use case with dozens of requirements. Coverage in the counties they operate in. Refresh cadence that lines up with their underwriting cycle. Field population where their model actually reads. A schema their engineers can ingest without a six-month project. Licensing terms their legal team will sign. Move any one of those and the others do not hold still. Deeper history often means staler records. Broader coverage often means thinner fields. There is no axis. There is a surface, and every buyer stands at a different point on it.
The scalar and the vector
This is the whole disease in one sentence: "better" is a scalar imposed on a vector. A scalar is a single number. A vector is a list of numbers pointing in a direction. The product is a vector. The buyer is a vector. The fit between them is the angle between two vectors, and an angle cannot be reported as "higher." It can only be reported as aligned, or not, and aligned with respect to what.
When you flatten a vector to a scalar, you throw away every dimension but one. That is not a rhetorical flourish; it is literally what the word does. "Better" is lossy compression, and the thing it compresses away is the only thing that decides the deal: which dimension mattered to this buyer. The rep who says "we are better" has quietly picked the axis for the customer and then congratulated himself for winning a race the customer never entered.
The gas station
Two pumps, both in spec, and not the same fuel.
The clean way to see this is a gas station, because gasoline is the closest thing to a true commodity most people buy, and even it is not one.
Pull into a Chevron and a BP on the same corner. Both sell 87-octane regular. Both meet the ASTM spec. By every rule that makes gasoline "gasoline," they are the same product, and the attendant would tell you so, and the attendant would be right at the resolution he can perceive. But they are not chemically identical, and no one in the industry pretends they are. They come from different crude slates, run through different refineries, and, most of all, carry different detergent additive packages. The TOP TIER standard exists precisely because that within-spec variation is real enough to gum up an intake valve over sixty thousand miles. Two fuels, both fully in spec, measurably different in the one dimension, engine deposits, that a spec sheet does not grade.
Now watch the word. Is Chevron "better" than BP? The pump attendant cannot answer; he cannot perceive the difference at all. And here is the part people miss: the chemist would not say "better" either. The chemist would say the additive package is different, and then would ask what engine, driven how, for how long, before venturing whether the difference helps you. The more you actually know, the less you reach for "better," because expertise is precisely the ability to see the vector that the layman has flattened into a scalar. "Better" is what you say when you cannot see the dimensions. It is a confession of low resolution dressed up as a verdict.
Fit is a property of the match, not the product
The honest word is fit, and fit has a grammar that "better" does not. Fit is not a property a product carries around in its pocket. It is a property of the relationship between a product and a use. The same dataset that is a catastrophic fit for a national insurer pricing wildfire risk is a beautiful fit for a regional lender checking clear title. Nothing about the data changed. The buyer changed, and fit is measured across the join, not on either side alone.
This is why "fit not features" is not a softer, humbler way of saying "better." It is a different claim about where value lives. Features live in the product. Fit lives in the space between the product and the problem, and that space is the only place a sale actually happens. When I stop selling "better" and start selling fit, I stop making a claim I cannot honestly defend and start doing the one thing that is genuinely useful: locating, with the buyer, the exact point on the surface where their requirements and my product coincide, and being honest about where they do not.
The buyer feels the difference immediately, because the buyer has been lied to by "better" a thousand times and has developed antibodies. Everyone says they are better. It is noise. The moment you say, "for what you are doing, here is where we fit cleanly and here is where we do not," you have said the first true thing they have heard all quarter, and trust is nothing more exotic than the accumulated evidence that you will keep saying true things.
The same disease everywhere
Once you can see the scalar-stapled-onto-a-vector move, you see it running underneath most of the things that make modern selling feel hollow.
It is the disease of the activity metric: flatten a rich, multidimensional quarter of a salesperson’s judgment into calls-dialed and emails-sent, then manage the scalar and wonder why the vector rotted.
It is the disease of the CRM stage-gate: flatten a living, path-dependent negotiation into a single progress bar from 0 to 100, then hold the number accountable and let the actual deal quietly diverge from it.
It is the disease of the data bake-off: flatten three vendors, each strong on a different axis, into one leaderboard, declare a winner, and buy the product that scored highest on the dimensions the test happened to measure, which are never the dimensions that break you in production.
Every one of these is the same error: a vector, violently compressed to a scalar, and then the scalar mistaken for the thing. "Better" is just the retail-facing version, the one the salesperson holds in his mouth. It is the smallest unit of the lie, and because it is small and constant and expected, it is the one nobody thinks to question.
Why this matters for the work
I am not making an argument for humility. I am making an argument for resolution. The reason I will not say "better" is not modesty; it is that the word is imprecise, and imprecision loses deals to people who cannot even articulate why they went the other way. The buyer who chose the competitor did not choose a "worse" product. They stood at a different point on the surface, and the competitor fit that point, and my rep never found out where they were standing because he was busy winning a footrace on an axis of his own invention.
So I have made it a personal rule, and it is the closest thing I have to a philosophy of the job. Never say better. Say fit, and then earn the right to say it by doing the actual work of finding where it holds. It is slower. It forfeits the easy dopamine of the superiority claim. And it is the only version of selling that is both honest and, in anything genuinely complex, the one that wins.