Why businesses end up running a dozen disconnected tools (and what integration involves)

The short answer

Nobody sets out to run a dozen systems. A team has a problem, buys a tool that solves it, and moves on; then another team does the same. Each decision is reasonable. The result is a business where the same customer, job or pupil exists in several places, and people spend part of every week keeping those copies in step.

Integration means deciding, for each piece of information, where it lives and how it reaches the systems that need it. The technology is rarely the hard part. The decisions are — and they are worth making before anything is built or bought.

How the tool count grows

The pattern is familiar from almost every organisation we work with:

• Sales buys a CRM. Finance already has accounting software.
• Operations adopts a job or scheduling tool, because the CRM does not do scheduling.
• Someone builds a spreadsheet to track what none of the systems cover.
• HR, support, compliance and marketing each add their own.

None of these choices is wrong. The problem is that each tool keeps its own copy of the things the business actually runs on — customers, jobs, people, money — and nothing keeps those copies in agreement except the people typing into them.

Why the connections are the problem, in arithmetic

If every tool has to talk to every other tool directly, the number of possible connections grows much faster than the number of tools. For n tools it is n × (n − 1) ÷ 2:

• 5 tools: 10 possible connections
• 8 tools: 28
• 12 tools: 66

Twelve tools connected directly can mean up to 66 separate links, each able to break on its own. Connected through one hub, the same twelve need 12.

Not every pair needs connecting, so real numbers are lower. But the shape explains why businesses that connect tools one pair at a time end up with a fragile web nobody fully understands — and why the alternative, a single integration hub, gets more attractive with every tool you add.

What the gaps cost

Two costs, and only one of them shows up in timesheets.

Re-keying. Take an illustrative business making 30 customer or record changes a week — a new customer, a changed address, a new job — each copied into 4 systems at 3 minutes a time. That is 30 × 4 × 3 = 360 minutes, or 6 hours a week. Multiply by 4.33 for a month (about 26 hours) and divide by 162.5, the hours in one UK full-time month: about 0.16 of a person.

Disagreement. The larger cost is when copies drift. Two systems show different addresses, different prices or different dates, and someone has to find out which is right — often after a customer has noticed. This cost rarely gets measured, because it appears as a query, a correction or a credit note rather than as time spent integrating.

What integration actually involves

1. Choose the master copy. For each thing the business runs on — customer, job, invoice, pupil, tenancy — decide which system owns it. Everything else reads from that one.
2. Map the fields. A "customer" in the CRM and a "contact" in accounting are rarely identical. Someone has to decide how each field corresponds, and what happens when one system has a field the other lacks.
3. Decide the timing. Instantly, as changes happen, or on a schedule. Instant is better for things people act on; scheduled is often fine for reporting.
4. Plan for failure. When a system is down, or two records conflict, what happens — and who finds out? An integration without alerts fails silently, which is worse than no integration.
5. Keep a record. What moved, when, and whether it succeeded, so problems can be traced — see audit trail.

Steps one and two are business decisions, not technical ones. They are also where most integration projects go wrong: the code is written before anyone has agreed which system is right.

Connect, consolidate or replace

Connect when the tools each do their job well and the pain is in the re-typing between them. This keeps what works and is usually the cheapest meaningful fix. A shared data layer and workflow automation do most of the work.

Consolidate when several tools overlap — two places that store customers, three that send messages. Retiring one is often simpler than connecting all three.

Replace when a tool is the bottleneck itself, cannot share its data, or is being worked around more than used. We cover how to tell in bespoke CRM vs off-the-shelf.

The same pattern runs through our sector articles on lettings agencies, colleges and charities: the systems are usually fine, and the seams are where the hours go.

Start with a map, not a build

Before connecting anything, draw the map. List every tool, what each one stores, and who types into it. Then mark every place the same information is entered twice. That page is the integration plan; everything else is implementation.

If the map shows three or four tools and little duplication, you may not need integration at all. If it shows the same customer in six places, measure the re-keying with the method in how many hours automation can save, and start with the seam that costs the most.