Your Data Isn't 80% of the Problem: Tracing a Myth
The "80% of AI is data prep" figure comes from a 2016 survey of about 80 people, and it is two numbers added together. Here is the trace.
20 notes on operational-efficiency.
All · manufacturing · ai-automation · field-sales · operational-efficiency · roi · industry-trends · sales-productivity · benchmarks · sales-operations · crm · quoting · procurement
The "80% of AI is data prep" figure comes from a 2016 survey of about 80 people, and it is two numbers added together. Here is the trace.
The UK's only official AI buying guidance is from 2020 and written for government. Nine contract terms the ICO makes mandatory, and a badge that misleads.
I counted 2,814 app store reviews of 14 field service platforms. Buyers average 3.33 stars; the engineers made to use it average 1.53. This is that gap.
The cost of error in UK construction is £210,000 per £1m of project value — and £160,000 of it never reaches a cost report. Traced to the primary source.
A pallet moves once; its paperwork gets handled five times. Mapping the admin at each stage of a UK 3PL workflow — and the two places a model must not go.
UK workers lose 31% of the week to admin. This index prices it role by role on ONS ASHE pay: hours lost, cost per person per year, and the gaps I won't fill.
What document extraction actually costs (£1.13 per 1,000 pages for plain OCR), what accuracy to expect by document type, and why no vendor publishes one.
Manufacturing investment fell 7.1% in the year to Q1 2026 while whole-economy investment rose. UK industrial electricity costs 83% more than the European median. Forty statistics on where the productivity gap actually sits, every one linked to a primary source and date-stamped, with a methodology section so you can check the workings.
Bain 2025: without process redesign, companies automate inefficiencies instead of removing them. If your quoting process is broken, AI quotes broken faster. Why 70% of AI projects fail and the five patterns that explain it.
Same trucks. Same drivers. 25% more deliveries. That's what SIG Plc got from AI-driven route optimisation. A practical guide to the three highest-value AI applications for UK distribution and wholesale companies with 50-250 staff.
Field sales companies across UK manufacturing, construction, and logistics are quietly breaking. Not failing — just getting more expensive every year. The 62% admin tax, the hiring trap, and the three gaps throttling revenue.
Somewhere in your company there's a spreadsheet that shouldn't exist. It was created because two systems don't talk to each other. That's a Ghost Workflow. How to find them, count the cost, and eliminate them.
Your pipeline says £2.4M in Q2. Your gut says half that. Your gut is probably closer. Why the data feeding your field sales forecast is fiction — and how signal-based forecasting fixes it.
When the sale is made, details die between field and factory. The non-standard finish, the adjusted delivery window, the revised spec. How UK manufacturers lose 3-7% of revenue at the handoff — and how to fix it.
Most UK manufacturers don't realise how much manual processes cost. We've broken down the hidden waste: quoting delays, CRM admin, handoffs, exceptions. See the numbers.
CRM adoption fails in field sales not because reps are lazy — the tool was built for desk-based teams. 79% of opportunity data never enters the system. Here's what UK manufacturers and field sales companies are doing instead.
When UK SMEs can't adopt AI, the consequences go beyond individual firms. Productivity bifurcation, supply chain fragility, and labour stagnation are creating a hollow middle in the UK economy — what it means and what needs to change.
AI adoption barriers aren't the same across sectors. Manufacturing hits infrastructure and data walls. Logistics runs into an ROI paradox at SME scale. Professional services gets caught in the billable hour trap. A sector-by-sector breakdown of why AI projects fail, and what to do about it.
UK field sales reps sell for 12 hours a week. The other 28 disappear into admin, CRM entry, quoting, and meetings. 27 sourced data points from Salesforce, SPOTIO, Gartner, Validity, and Forrester showing exactly where the time goes — and what it costs.
The gap between enterprise AI adoption and UK mid-market businesses is wider than most people admit — and it's compounding every month. Data from Salesforce, ONS, Eurostat, OECD, and Anthropic on where the real adoption numbers sit.