Build a B2B social selling benchmark range, then expose its weakness

What to take away

No authoritative England-wide B2B social selling benchmark exists, so build an internal reference range instead.
Fix the numerator, denominator and cohort rule before you look at any percentage.
ONS register data can define an English business population but cannot measure replies or buying intent.
Report sample size, median and a stated percentile range, plus what would invalidate it.
Treat an out-of-range result as a review trigger, not proof of poor work.

Write the measure before you collect the data

Name the decision the range will support: falling reply quality, a comparison between account groups, or a resourcing trigger. One figure cannot serve all three. So write the decision down first.

Numeratorunique delivered contacts who give a relevant reply within 30 days.
Denominatorunique contacts receiving an eligible approach in the entry month.
Exclude verified delivery failures and non-marketing service replies.
Retain silence and refusals, and state whether a referral qualifies.

Keep the definition identical across the reference cohort and the current one, or the comparison breaks.

Before fixing a reply rule, read Five ways to attribute B2B social selling outcomes, because each method proves something different about the same reply.

What ONS and supplier sources can and cannot supply

The ONS UK business activity, size and location dataset counts registered enterprises and local units by geography and business characteristics. Enterprise location is usually the main operating site or head office.

That snapshot cannot measure platform use, messages, replies or buying intent. Match the statistical unit to your account rule, and note businesses outside the VAT and PAYE register.

What it measures

ONS business activity, size and location: Registered enterprises and local units by geography and characteristics
LinkedIn Sales Navigator metric guide: Product usage and effectiveness events for stated subscriptions
Your own CRM extract: Eligible approaches and replies under your rule

What it cannot give you

ONS business activity, size and location: Replies, platform use or buying intent
LinkedIn Sales Navigator metric guide: Your internal outcome rate
Your own CRM extract: An England-wide comparison

LinkedIn's Sales Navigator metric guide defines product usage and effectiveness events and limits reporting to stated subscriptions. A supplier event may differ from your rule, and opted-out or unmatched activity alters coverage. Do not splice product-defined activity into an internal outcome rate.

A supplier engagement rate and an internal reply rate are different measures wearing the same word.

Build a comparable cohort

Choose records before viewing results, then match what moves the outcome:

Choose records before viewing their results.
Match account eligibility, business size band, sector and English geography rule.
Match seller role, channel, contact type and observation window.
Record the inclusion query, extraction date and duplicate treatment.
Report the missing-data sharerecords lacking a comparison field divided by all eligible records.

Exclude authorised tests and state the period. Data on one basis is comparable; separately collected data needs closely aligned definitions and methods.

Example: a substantive-response reference range

The figures below are an illustration, invented for arithmetic and not drawn from a published source.

Take 24 seller-periods across six England-registered software firms, each covering at least 150 eligible approaches. The 24 reply rates run from 3% to 26%.

Statistic / Value

Seller-periods in reference cohort: 24
Median substantive reply rate: 11%
10th percentile: 5%
90th percentile: 20%
Current period: 3.3%, from 5 replies to 150 approaches

The current period sits below the 10th percentile, but one extra reply would add about 0.7 percentage points, so the shortfall is fragile.

Where the reference range breaks down

Preserve individual seller-period values, then report sample size, median and a stated percentile range. Never drop awkward observations after seeing the distribution. If the mix or the measurement system changed, split or rebase with a documented reason.

The Infrastructure and Projects Authority's benchmarking guidance covers major projects, not sales. Its transferable method is to validate and rebase comparable historical observations, and to disclose how fragile the benchmark is.

Cohort rules built on assumption fail quietly, so test the account frame with account frames, interviews and surveys before it becomes the reference set.

What to do when performance falls outside the range

Check source completeness, cohort mix and operating changes before concluding that outcomes moved. Record the decision, the evidence sought and the rebuild date.

If the range drives spending, a B2B social selling budget template keeps supplier cash separate from staff time, so the trigger lands on the right line.

Common questions

Why is there no authoritative England-wide benchmark?

The sources reviewed on 5 September 2026 contain no England-wide performance benchmark for B2B social selling. ONS data covers registered enterprises and local units, not platform use, messages, replies or buying intent.

How do I define the numerator and denominator?

The numerator is unique delivered contacts giving a relevant reply within 30 days. The denominator is unique contacts receiving an eligible approach in the entry month. Exclude delivery failures and service replies, retain silence and refusals, and say whether a referral qualifies.

What if current performance falls outside the range?

Check source completeness, cohort mix and operating changes before concluding that outcomes moved. One extra reply can shift a small sample by around 0.7 percentage points.

How small can a cohort be before the range is meaningless?

Below about 20 seller-periods the percentiles move too much to set a target. Report the count and the spread instead, and keep the values descriptive.