A better specification does not sell an instrument
A better specification rarely wins a laboratory sale on its own. The lab already owns a working machine with an approved method attached to it, and replacing that machine costs validation time, retraining and risk that no specification line item pays back. Winning means finding value where your advantage is worth that cost.
The specification war nobody wins
Almost every instrumentation startup I have worked with arrives with a comparison table. They claim their specification is better on multiple levels, be that sensitivity, throughput, or range. The conclusion drawn from that table is that the market has not yet heard about them, and the fix is more awareness.
Usually the market has heard, evaluated, and decided that those improvements were not worth the disruption.
Consider what actually happens when a laboratory replaces an instrument. The method that runs on it has been validated. Someone wrote that validation, someone approved it, and in a regulated environment someone will have to do both again. Analysts trained on the old workflow need retraining. Historical data now sits behind a method bridge. There is a service contract to unwind and a new one to negotiate. If the instrument sits in a release-testing path, the risk of a delay has a number attached to it that dwarfs the purchase price. And there is a quieter risk on top of all that: a better instrument may detect things about their product that were previously invisible, which in a regulated environment can be catastrophic.
Against all of that, your resolution improvement has to do more than exist. It has to solve something the lab is currently blocked on.
Blocked, not merely improvable
The useful question is not "is our instrument better?" but "who is currently unable to do their job with what they have?" Those are different populations, and the second one is much smaller and much easier to sell to.
A laboratory that cannot resolve a bimodal size distribution it knows is there will move. A laboratory whose current single-mode result is adequate for release will not, however much better your figure is. Same instrument, same specification, two completely different sales.
This is why segment selection does more work than messaging. Get the segment right and the message writes itself out of the buyer's own description of the problem. Get it wrong and no amount of language rescues it.
What to do instead
Start with the people who evaluated you and chose someone else. They are more useful than your customers, because they have already articulated the reason not to buy, and they usually do it in operational terms: revalidation, headcount, timing against a filing.
Then take your comparison table and mark each row with the name of a laboratory for whom that row is a blocker. Rows with no name against them are footnotes, not claims. The rows with names are your market.
None of this makes the specification irrelevant. It moves it from the headline to the proof, which is where a technical buyer expects to find it anyway.
Written by Matthew McGann, President and Owner of Particle Analytx. More about the practice.
Questions on this piece
Does specification never matter?
It matters as proof, once a buyer already wants to switch. It rarely creates the desire to switch on its own.
How do I find the segment where my advantage matters?
Interview people who evaluated you and bought something else. They will describe the switching cost precisely, because they paid attention to it.
What if my advantage is genuinely large?
Then say it, with the method attached — and still identify who is currently constrained by the thing you improved. A tenfold improvement in something nobody is blocked on is still a hard sell.
Recognise the problem?
If this describes where you are, thirty minutes on a call will get further than another article.