Research Methodology
LeadLinka Research ยท Last updated: 24 June 2026
LeadLinka Research applies one consistent methodology to every figure we publish. We do not improvise keyword sets, date fields, or comparison windows from article to article. If a figure cannot be reproduced from the definitions on this page, we do not publish it. This page covers the project-type topic categories used in our insights (data centre, battery storage, EV charging, and grid).
1. Population and source
- Source: every figure is drawn from the public planning-application records published by UK local planning authorities, which LeadLinka tracks across roughly 1.68 million applications.
- Matching: we match on the text of each application's description using a case-insensitive search.
- Date basis: every period is cut on the application's validation date (the date the authority validated the application). This is the closest available proxy for when demand was registered. We do not use decision dates, which measure outcomes rather than demand and are recorded less consistently.
- No status filter: we count every application matching the keyword in the window, regardless of its later status or decision. We are measuring applications submitted, not approvals. A withdrawn or refused application still represents real demand activity at the point of submission. Where an approvals-only figure is needed, we define it explicitly and separately.
- Missing dates: an application with no validation date is excluded from every window, because it cannot be placed in time. This is intentional.
Approval rate definition
Where LeadLinka Research publishes a council approval rate, it is calculated as approved decisions divided by total decided applications in the period (decided-only basis; pending and withdrawn applications are excluded). The approved flag is the value recorded by the council in its own planning system. For some authorities this includes consultation responses, permitted-development determinations, and certificate-of-lawfulness outcomes alongside conventional full and outline decisions. We apply the council's own flag consistently and do not re-classify decisions. This means approval rates are comparable across time for a given council, but cross-council comparisons should account for the fact that some authorities record a wider range of decision types than others.
2. Category keyword definitions
These are the exact match expressions. They are deliberately narrow to moderate to keep precision, so they undercount relative to a broad "any mention" set. That is an intended, defensible trade-off.
| Category | Description matches |
|---|
| Data centre | data centre or data center |
| BESS / battery storage | battery (energy) storage, or the whole word BESS |
| EV charging | EV charg, electric vehicle charg, or rapid charg |
| Substation / grid | substation or grid connection |
Known limitations, stated openly:
- These count applications whose description mentions the term, not a verified project-type classification. We phrase results as "applications mentioning" or "applications for" a category, not as "[category] projects".
- The substation and EV-charging terms can catch incidental mentions, for example a substation or charging bay inside a larger housing scheme. This inflates the absolute counts, but it applies consistently across periods, so the trend percentages remain valid.
- We hold a separate, smaller curated classification internally, but it is a precomputed subset, not a market denominator, and we do not use it for these counts.
3. Comparison windows
Windows we use
- Q1-over-Q1 year-on-year (for example Q1 2025 versus Q1 2026): a complete quarter in both years. This is our preferred headline window.
- Rolling 12 months versus the prior 12 months: a larger sample with less seasonal noise.
Windows we do not use
- Calendar-year year-on-year (2024 versus 2025): we do not use this. Our council coverage grew through 2025 as we onboarded toward roughly 296 councils, so calendar-year counts are inflated by more councils being tracked, not by more demand. Every category shows a spurious increase of +57% to +76% on this window, so it is not a demand signal. We exclude it deliberately, and we think saying so is the honest thing to do.
- Any window ending in an incomplete period: the current quarter is partial before it ends, and data lags slightly from collection and validation. Comparing a partial period against a complete one manufactures a false decline, so we never do it.
Robustness check. For any headline percentage we also compute it on a fixed council panel: councils that had at least one application in both periods. If the panel figure is close to the all-council figure, the coverage confound is not biting and the figure is safe. In practice, our Q1-year-on-year and rolling-12-month figures move by less than 3 points under the panel, while calendar-year figures collapse.
These are our live, defensible numbers, verified on 24 June 2026 using a like-for-like council panel. We refresh them on the same definition as the data updates.
| Category | Q1 2025 to Q1 2026 (panel n=257) | Rolling 12 months (panel n=270) |
|---|
| Data centre | +32% (28 to 37) | +4% (140 to 146) |
| BESS | -54% (204 to 94) | -31% (739 to 507) |
| EV charging | +17% (510 to 596) | +8% (2,037 to 2,204) |
| Substation / grid | -6% (442 to 416) | -1% (1,827 to 1,806) |
Absolute population and estimated construction-cost pipeline (all councils):
- Data centre: 343 applications, £23.99bn to £46.89bn
- BESS: 1,337 applications, £1.06bn to £2.49bn
- EV charging: 5,212 applications, £4.05bn to £7.07bn
- Substation / grid: 4,678 applications, £29.55bn to £57.02bn
Pipeline ranges are indicative estimates based on UK benchmark build rates, not quotes or a guarantee. Counts reflect applications mentioning each category, as defined in section 2.
5. How to cite our figures
Every published figure carries four things: the number, the comparison window, the source, and a method note. For example:
"EV charging planning applications grew 17% year-on-year in Q1 2026 (like-for-like council panel), according to LeadLinka Research."
Short method note for footnotes:
LeadLinka Research. Applications validated in the period whose description references the category; like-for-like council panel; calendar-year comparisons excluded due to changing coverage.
Journalists and researchers are welcome to cite these figures with attribution to LeadLinka Research. For questions or a specific data cut, contact support@leadlinka.co.uk.