Why vendor-independence matters
Ask a vendor whether you need their product and you already know the answer. That is not a knock on their character; it is simply how incentives work. Their business depends on selling a platform, a licence, or a build, so the question you actually care about - do we need this at all? - is not the one they are set up to answer. In their hands, "should you do this?" quietly becomes "which of ours would you like?"
Most of the AI advice reaching a shipping business today comes from someone with a stake in the outcome. The software company wants a subscription. The systems integrator wants a large implementation. Even a well-meaning consultant is often reselling a particular tool under the bonnet. None of that makes the advice worthless. It does mean you are rarely hearing an opinion that is free to say no.
The incentive hiding in "free" advice
The trouble with conflicted advice is not that it is dishonest. It is that it is systematically tilted. A vendor's demo will always find a use case for the vendor's product. A tool built to summarise documents will recommend summarising more documents. Ask a company that sells large builds whether you should build, and the answer will lean towards building. The bias is rarely a lie; it is a gravitational pull, and over enough decisions it steers you towards spending more, integrating more, and depending more than your problem ever required.
For a shipping business, that pull is expensive in a way the headline price does not capture. You pay for the licence, yes, but you also pay in the six months your team spends bedding in a tool that solved a problem you did not really have, while the one that mattered went untouched.
What independence changes in practice
When the person advising you has nothing to sell but their judgement, the conversation changes shape. The first question stops being which tool? and becomes what, precisely, are we trying to fix, and is it worth fixing this way? Only after that does the discussion move to whether an off-the-shelf product, a custom build, or a change to your existing process is the right response.
That independence also lets us be straight about trade-offs. We can tell you that a well-known platform is a poor fit for your data setup, or that the impressive-looking tool your competitor bought would be overkill for your volumes, without worrying about a sales target. Our recommendation might be a third-party product we have no commercial relationship with. It might be that you already own most of what you need and simply are not using it well. The point is that the advice is free to follow the problem, not the invoice.
Sometimes the honest answer is "not yet", or "not at all"
The clearest sign of independent advice is that it is willing to talk you out of a purchase. Plenty of shipping problems are not AI problems. Some are data problems that a tool would only paper over; some are workflow problems that a conversation and a spreadsheet would fix faster and more cheaply. A vendor is not well placed to tell you that. We are, because saying "you do not need this yet" costs us nothing but earns your trust and, we would argue, that is the more valuable thing to offer.
This is not anti-technology. We use AI heavily in our own work and we are enthusiastic about where it genuinely helps. It is simply that enthusiasm and independence are not in tension. The best way to be trusted on where AI does pay off is to be equally honest about where it does not.
How to tell whether your adviser is truly independent
If you take one thing from this, let it be a short set of questions to put to anyone advising you on AI:
Ask who pays them. If their income depends on you buying a specific product, factor that into every recommendation they make. It does not disqualify them; it just needs naming.
Watch how they treat "no". An adviser who never concludes that the answer is don't is not weighing your case; they are steering it.
Notice which comes first, the problem or the product. Independent advice starts with your operations and works towards a tool. Conflicted advice starts with the tool and works backwards to a justification.
Check whether the recommendation could be a competitor's. Someone truly independent will happily point you to a rival product, or to no product at all, if it serves you better.
Vendor-independence is not a slogan for us - it is the whole reason that Libra AI exists. We are business consultants, not a software company, and we are not paid to over-build or to push a particular tool. That is precisely what lets us give you an answer you can trust including, when it is the right one, the answer that saves you the spend.
If you would like an independent view on where AI fits in your business, let's have a chat.
