There is a cost in your business that never shows up on an invoice, never appears as a line in your accounting software, and never gets discussed in a budget meeting. It is the time lost every day to information that does not flow: the number rekeyed into a third system, the update that lived in someone’s head, the customer who slipped through because the booking tool and the payment tool were never introduced to each other. Call it the communication tax. Most businesses pay it every day, and almost none of them have ever seen the bill.
This piece is the full argument, and it is deliberately a longer read than most business writing on the subject, because the difference between noticing a problem and fixing one is the level of detail at which you understand it. It defines the tax precisely, shows where the concept comes from in my doctoral research, explains why the cost compounds instead of staying flat, gives you a formula to price it in your own business, takes the strongest objections seriously, and ends with a one week audit that will put a defensible dollar figure on it by Friday.
What the communication tax is
A definition worth using has to separate the tax from the work. Talking with customers is not the tax. Selling, scheduling, diagnosing, and delivering are not the tax. Those are the business. The communication tax is everything a person has to do to move information the business already possesses from where it is to where it is needed. The moment a human being becomes the transport layer between two systems, two people, or two moments in time, the meter is running.
Three properties make this cost uniquely hard to see. First, it is charged per movement, not per month. There is no subscription line for it, so no report surfaces it. Second, each individual charge is trivial. Ninety seconds to rekey a phone number does not feel like a business problem, and nobody escalates ninety seconds. Third, the people paying it are usually proud of it. Chasing the update, catching the dropped handoff, and holding the whole picture in your head all read as diligence. The tax disguises itself as conscientiousness, which is why conscientious teams pay the most.
The four line items
Priced or not, the tax always arrives in one of four forms. Naming them matters, because the audit at the end of this piece asks you to log a week of incidents against these categories.
Translation is the same information re-entered or reformatted because two systems do not share it. The estimate copied from the quoting tool into the invoice, the customer details typed into the scheduler and again into the waiver form, the job notes photographed on paper and then summarized into the CRM. Every translation event is a person doing an integration’s job.
Chasing is status that has to be pulled instead of pushed. Where are we on the Hendricks job, did the deposit come through, has anyone called the supplier back. Chasing consumes two people per incident, the one who asks and the one who stops working to answer, and it multiplies in businesses where the owner is the only person who can see across the whole operation.
Drops are handoffs that fail silently. The lead that came in through the website and was never called, the reschedule that made it to the calendar but not to the technician, the intake form completed on paper and filed instead of entered. Drops are the most expensive line item because they are priced in revenue rather than minutes, and the business usually never learns the drop occurred.
Rebuilding is context reconstructed because it was not captured where the work happens. The meeting that exists to re-establish what was already decided, the twenty minutes spent searching an inbox for an attachment, the new employee who has to interrupt someone because nothing is written where the answer should live.
Keep those four names in mind. They come back at the end.
What the research found
The communication tax is not a metaphor invented for a blog post. It is the pattern that sat underneath my doctoral research, and the reason I trust it enough to build a practice on it is the shape of the study that surfaced it.
My 2014 dissertation was a qualitative descriptive multiple-case study built on in-depth interviews with 19 leaders across Fortune 500 organizations, deliberately drawn from both sides of a technology decision: organizations that had adopted social business platforms and organizations that had resisted them (Goodwin, 2014). The side by side design was the point. Studying adopters alone tells you what satisfied customers say. Studying adopters and resisters against each other isolates the conditions that separate a decision from a delay, and it lets the contrast between the groups carry the findings rather than any one leader’s enthusiasm.
Eight themes emerged from those interviews. The one this piece is built on is that platform users standardize communication. The leaders who adopted largely did so because they had recognized that their communication was fragmented across too many disconnected tools and channels, and they saw a unified system as the fix. Recognizing a communication challenge was one of the conditions most associated with adoption. The finding that has stayed with me longest, though, is the inverse. The resisters often could not see the fragmentation at all. It had become the water they swam in. Adoption was frequently less about loving new technology and more about finally seeing a cost the organization had stopped noticing, and deciding to stop paying it.
The macro data from the same period says the cost those leaders were noticing was enormous. The McKinsey Global Institute’s 2012 study of what it called interaction workers, the managers and professionals whose jobs run on judgment and coordination, estimated that the average such worker spent 28 percent of the workweek managing email and nearly 20 percent more searching for internal information or tracking down colleagues who could help (McKinsey Global Institute, 2012). Nearly half the week, in other words, went to moving information rather than acting on it, and MGI estimated that fully implemented social technologies could raise interaction worker productivity by 20 to 25 percent. My interviews were the micro view of the same phenomenon: individual leaders either seeing, or failing to see, that their organizations were paying that bill.
The technology in question was social business platforms, and that matters for how you read the research today. The finding was never about a product category. It was about the mechanics of recognition: which leaders see an invisible cost, what allows them to see it, and what they do next. In 2026 the decision on the table is AI rather than a social platform, and the mechanics have not changed. The leaders moving first are, once again, the ones who can see the tax.
Why the tax compounds
A cost that stayed flat would eventually get noticed and killed. The communication tax survives because it grows along three curves at once, and none of them look alarming from inside the business.
The first curve is frequency. No single incident is worth escalating, but the incidents repeat, and repetition is where the money goes. A two minute rekeying event that happens eight times a day is not a two minute problem. It is over an hour a week, per person, for that one event type, and a working year multiplies whatever your week looks like by roughly two hundred and forty.
The second curve is team growth. Communication paths grow faster than headcount. Three people share three lines of communication. Six people share fifteen, and ten people share forty five. Every hire adds more new paths than the hire before, and every path is a place where information can stall, mutate, or die. This is why a business that ran smoothly at four people develops mysterious friction at eight, and why the owner’s instinct that things used to be simpler is not nostalgia, it is arithmetic.
The third curve is tool accumulation, and it is the cruelest of the three because it grows out of attempted fixes. Each tool added without an integration plan adds new places information can live without being shared, which means new translation events, new chasing, and new surfaces for drops. The modern evidence on this is blunt. McKinsey’s 2025 State of AI research found that, of all the changes organizations make, fundamentally redesigning workflows is the one most associated with real bottom line impact from AI, and that only about one in five organizations using generative AI has done it (McKinsey, 2025). BCG describes the same failure from the other side as a widening gap: companies accumulating scattered, unscalable proofs of concept that never integrate into how work flows (BCG, 2025). A decade of extraordinary software did not kill the communication tax, because most companies bought tools instead of redesigning flow. AI changes the economics, not because it is smarter software, but because the specific work that generates the tax, moving, reformatting, routing, and summarizing information, is exactly the work a well-built system can now do on its own.
Pricing the tax
An unpriced problem is an irritation. A priced problem is a capital allocation decision. Here is the working convention Omnine uses to put a first number on the tax, built to be conservative enough to defend in front of your accountant.
Annual communication tax = People, times minutes lost per person per day, divided by 60, times loaded hourly cost, times working days.
Take a six person service business. Assume each person loses 25 minutes a day to translation, chasing, drops, and rebuilding combined. Against the MGI finding that interaction workers spend nearly half the week on email and information hunting, 25 minutes, about five percent of a workday, is a deliberately gentle assumption. Use a blended loaded cost of 38 dollars an hour and 240 working days. Six people times 25 minutes is 150 minutes a day, which is 2.5 hours. At 38 dollars an hour that is 95 dollars a day, and across 240 working days it comes to 22,800 dollars a year.
Omnine estimates that a real audit almost always produces a larger number than this convention, for three reasons the formula deliberately excludes. It prices the owner’s minutes at the blended rate, when the owner’s time is worth a multiple of it. It counts zero dollars for drops, when a single dropped lead in a service business can be worth more than a month of everyone’s rekeying. And it ignores error cost entirely, even though every manual translation event carries a defect rate, and every defect costs more time downstream than the original entry did. The convention is a floor. Its job is to turn the conversation from whether the tax exists into what removing it is worth.
The tax in the field
Advocate Roofing, a residential roofing company in Etowah County, Alabama, came to Omnine with what looked like broken software. After a version migration in their quoting platform, estimate templates had lost their connection to the supplier’s synced price library. Every quote the sales team produced came out without materials attached, which meant every quote was rebuilt by hand before it could go to a homeowner. Worth pausing on: within weeks, the team had absorbed that rebuilding into their routine. A serious communication failure between two systems had already been normalized into everyone’s workflow, which is exactly the pattern the resisting organizations in my research displayed.
The diagnosis took an afternoon of systematic work once the problem was framed correctly. The software was not broken. The two systems had stopped speaking the same language, literally: product names in the estimate templates no longer matched the names in the supplier library they were supposed to pull from, and a second, redundant supplier account was interfering with the sync underneath. The fix was to repair the naming layer so the systems recognized each other again. Estimates then generated correctly, labor and materials included, and the sales team stopped being the integration layer between their own tools.
Lucky Little Riders, the youth riding school featured in the Omnine results, shows the other direction: removing the tax at the point of sale before it accumulates. Booking, intake, waivers, and payment now run as a single flow, so a parent books once and everything else follows without anyone rekeying anything. In both cases the deliverable was the same thing at different scales. Information the business already had began arriving where it was needed without a person carrying it.
The strongest objections
A concept this convenient to a consultant deserves adversarial review, so here are the three strongest objections I hear, taken seriously.
The first objection: this is just the cost of doing business, and every company has overhead. The research answer is that this sentence is not a rebuttal of the finding. It is the finding. Classifying the fragmentation as normal is precisely what the resisting organizations in my study did, and it is the mechanism that keeps the tax invisible. The adopters were not different because they tolerated less overhead. They were different because they had stopped filing an avoidable cost under overhead. Whether the tax is fixed or optional is not a matter of opinion anymore; the businesses that removed it settled the question.
The second objection: we will fix it with another tool. Sometimes that is true, and often it produces the opposite. A tool adopted without a flow redesign adds surfaces: one more login, one more place information can live without being shared, one more translation event per job. That is the pattern BCG’s widening gap describes at enterprise scale, and the pattern I wrote about at small business scale in why buying more AI tools is making you slower. The question that separates a fix from a new tax is never what the tool can do. It is which movements of information the tool eliminates end to end.
The third objection: we are too small for this to matter. Small businesses pay a higher rate, not a lower one. In a six person company every person is a single point of failure with nobody behind them, and the most expensive courier in the building is usually the owner, whose judgment is the scarcest resource the business has and who spends a distressing share of it forwarding, rekeying, and chasing. The dollar figure scales down with headcount. The percentage of capacity lost does not, and the strategic cost of the owner’s attention arguably scales up.
The one week audit
Everything above is only useful if it changes what you do on Monday, so here is the instrument. It costs one week of light attention and thirty minutes on Friday.
From Monday to Friday, everyone on the team keeps a running note with the four line items as headings: Translated, Chased, Dropped, Rebuilt. Every time an incident happens, they add one line recording what it was and roughly how many minutes it took. A rekeyed customer record goes under Translated at two minutes. A where are we on this message goes under Chased at five minutes, because it cost two people. The rules are strict on one point: observe, do not fix. The week is for measurement, and teams that start fixing mid week stop logging.
On Friday, spend thirty minutes. Tally the minutes by category and by incident type, then run the formula from the pricing section on your real numbers instead of my assumptions. Then rank the incident types by total minutes and apply the selection rule: your first system to build is the highest minute item that occurs daily and crosses at least two tools. The right target is not the most irritating item but the most expensive one. Irritation is a poor prioritization signal because it correlates with novelty, while expense correlates with repetition, and repetition is what automation is for. That selection discipline, starting from the problem rather than the technology, is the subject of AI adoption starts with a problem, not a tool, and the wider set of conditions that separate adopters from resisters is mapped in the 5 drivers of AI adoption.
Where this leads
The through line from the research to the field work is a single sentence: adoption begins when a leader finally sees a cost the business had stopped noticing. The audit gives you the sight, the formula gives you the number, and the number turns a vague sense that things are harder than they should be into a decision made on purpose. Some businesses will run the audit, price the tax, and rationally decide to keep paying it for now. That is a legitimate outcome. The only illegitimate outcome is continuing to pay a bill you have never read.
If you want the shortcut to seeing where your largest leak is, the Omnine AI Readiness Assessment takes about three minutes and maps your answers to the same framework this piece is built on.
References
Goodwin, M. R. (2014). A qualitative descriptive multiple-case study: Fortune 500 leaders’ social business platform adoption (Doctoral dissertation, University of Phoenix). ProQuest Dissertations Publishing (UMI No. 3648813).
McKinsey Global Institute. (2012). The social economy: Unlocking value and productivity through social technologies.
McKinsey & Company. (2025). The state of AI in 2025: Agents, innovation, and transformation. McKinsey Global Survey on the state of AI.
Boston Consulting Group. (2025). Are you generating value from AI? The widening gap.