Your tools don't talk to each other — the integration tax of running a business on disconnected apps

No business sets out to build a disconnected mess. It happens one good decision at a time. You add a tool for invoicing because the spreadsheet broke. A separate one for the online store. WhatsApp for sales because that is where the customers already are. An accounting package because the bookkeeper asked for it. Each was the right call on the day. Then one morning you look up and your business runs on a dozen apps, none of which know the others exist — and the same customer, the same order, the same number lives in five places that quietly disagree.

The scale of this is not a hunch. The average organisation now runs roughly 897 applications, and only about 29 percent of them are integrated [1]. The rest sit as siloed islands, and 90 percent of IT leaders say those silos are actively creating business problems [1]. The bridge between the islands is almost always a human: someone exports a file here and re-types it there, copies a number from the store into the ledger, pastes an address from WhatsApp into the delivery sheet. That bridge is the integration tax, and you are paying it whether or not it appears on any invoice.

This piece is the operator's map of that tax — not "buy a bigger platform," because that is often just a more expensive island. The real questions are: what is the disconnection actually costing you, why does bolting on one more integration at a time make it worse rather than better, and what is the one structural decision that ends it.

Start with what the disconnection costs, because operators feel it long before they price it. The most-cited number is the toggle tax: a Harvard Business Review study that instrumented real users found people switch between applications and windows roughly 1,200 times a day, adding up to just under four hours a week spent re-orienting after each switch — about 9 percent of their time at work ^[2]. In one supply-chain task the researchers traced, a single transaction took about 350 switches across 22 different applications ^[2]. That is not a discipline problem you can train away; it is the direct, measurable cost of asking a human to be the integration layer between tools that were never connected.

The second cost is quieter and more dangerous: the numbers stop agreeing. When the same customer exists in the store, the CRM, the accounting tool, and a WhatsApp thread, there is no single source of truth — there are four, and they drift. An order marked shipped in one system is still "pending" in another. A price updated in the catalogue never reaches the invoice template. MuleSoft's benchmark puts hard figures on how widespread this is: 95 percent of organisations report integration challenges, and 80 percent say those challenges are what create their data silos in the first place ^[1]. You do not notice the drift until a decision gets made on the wrong version of a number — and by then it is a refund, a stockout, or a board slide that does not match the bank.

Now the structural reason it gets worse the more you patch it. Every time you connect two apps with a direct, point-to-point link, you add one more bridge to maintain — and the math is unforgiving. Five systems wired to each other is ten connections; ten systems is forty-five. Each link is hand-built against one vendor's export format and breaks the day either side changes an API or a column. This is why teams that "just add an integration" each time end up spending an average of 36 percent of IT effort on building and maintaining the connections themselves ^[1]. The wiring quietly becomes the system — a tangle nobody fully understands and nobody dares touch.

The way out is a change in topology, not another app. Instead of every tool talking to every other tool, you route them through one core system that owns the canonical record — the customer, the order, the invoice — and everything else reads from and writes to that. Five apps become five clean connections instead of ten brittle ones, and adding the sixth app is one new link, not five. The core does not have to be a giant platform; it has to be the one place where the truth lives, with the rest wired around it. That is the difference between owning a system and renting a pile of islands.

This bites hardest in exactly the markets Felukaa builds for. A growing MENA operation typically runs the most disconnected stack of all: a global SaaS store, a local payment gateway, WhatsApp as the busiest sales channel, an accounting package built for local tax, and at least one spreadsheet holding it together — tools from five vendors who will never integrate with each other on your behalf. With regional commerce expanding at double digits toward a market that could pass 57 billion dollars by 2029 ^[3], the businesses scaling into that growth on a bolt-on stack are scaling the integration tax with it. Smaller firms feel it sooner than they think: a company under 200 people already runs around 42 separate SaaS tools on average ^[4], which is more than enough islands to drown in.