The enemy here is not any one vendor. Most of the tools in a typical pharmacy software stack are good at the thing they were built for, and several of them are better at it than a suite ever will be. The problem is what happens in the space between them, which is nobody’s product and therefore nobody’s responsibility.
Here is a composite stack. It is not exaggerated, and if you run an independent you will recognise most of it.
- The dispensing system, which knows what went out the door.
- A reconciliation service, which knows what the PBM eventually paid.
- A purchasing or wholesaler tool, which knows what you bought and at what invoice price.
- A clinical services or patient engagement platform, which knows which vaccinations and reviews happened.
- A compliance or audit-support subscription, which knows what a PBM will want if it comes asking.
Five systems. Between them they hold every number needed to answer one question: did this pharmacy make money this week, and on what. None of them can answer it, because the answer requires joining all five.
Where the money actually goes
Swivel-chair time. Somebody exports a report from one system and pastes it into a spreadsheet next to an export from another. That person is usually the owner or the strongest technician, which means the cost is not their hourly rate, it is the dispensing or the patient conversation that did not happen while they did it. This work is invisible on every invoice and is frequently the single largest line item in the stack.
Data that never joins. The purchasing tool identifies a drug by NDC and package. The claim carries an NDC, a quantity dispensed and a days supply. The clinical platform records an encounter with a CPT code and no NDC at all. Nothing shares a key. So the join is done by hand, by eye, on a sample, once a month, if at all, and a sample is not a control.
Alert fatigue across five inboxes. Each tool is well-behaved on its own and each one emails you. None of them can tell you that the MAC change flagged by one is the same event as the underwater fill flagged by another and the purchasing shift recommended by a third. You get three notifications about one problem and treat them as three problems, or, more often, as noise.
Reconciliation lag as a structural blind spot. Most of the stack reports backwards. The purchasing tool tells you what you already bought. The reconciliation service tells you what you were already paid. By the time both have spoken, the decision they would have informed was made weeks ago and made forty more times since.
No single view of the week. This is the one owners describe last and feel first. There is no screen anywhere that says: this is what we dispensed, this is what it cost, this is what came back, this is what is still unbilled, this is what is disputed. Assembling that view is a project every time, so it happens quarterly at best, and a business that measures its margin quarterly is a business making purchasing decisions blind for eleven weeks out of twelve.
Three questions to audit your own stack
Ask them literally, and time yourself.
- How long does it take to answer “which NDCs lost money last month, on which plan”? If the answer involves more than one export, the join is the product you are missing.
- What did we do clinically last month that we never billed? If no system holds both the encounter and the claim, the gap is not measurable, which means it is not managed.
- If a PBM requested two hundred claims tomorrow, which system would you open first? If the honest answer is “several, and then a filing cabinet”, your audit readiness is a person, not a process.
None of this argues that consolidation is automatically right. A single vendor that is mediocre at five jobs is worse than five specialists, and anyone selling you a suite should have to prove each product would survive on its own. The argument is narrower: the joins between your tools are doing real work in your business, that work is currently being done by a human with a spreadsheet, and it should be counted as a cost of the stack rather than treated as free.