Measurement · ROI · attribution · cohort analysis · Armenia · B2B

How to Measure Merch ROI for B2B Programmes

The hardest part of a B2B merchandise programme is not the sourcing, the production or the rollout, it is the question that arrives at the end of the year: what did we actually get back from what we spent. Most teams answer that question with a number that satisfies the board and misleads the next budget, because the cost side is usually undercounted — nobody adds the graphic-designer hour, the warehouse labour or the freight revision — and the return side is usually overcounted, because a recipient smile is real evidence in the moment and no evidence twelve months later. This page is for marketing, brand, procurement and finance leads in Armenia and the wider EAEU who want a measurement approach that survives an audit, not a measurement approach that survives a single review meeting.

Request a measurement briefAll B2B how-tos

Why most merch ROI numbers fail on first contact with finance

A B2B merchandise programme produces data on both sides of the equation, but most annual reviews report only the side the team measures easily. The cost side arrives at the programme manager in twelve to twenty invoice lines per order and is then collapsed to the order total in the programme budget. The return side arrives as anecdote, photo and survey and is then collapsed to a single qualitative claim in the board pack. The collapse is where the measurement breaks: the cost figure survives because it is the invoice total, and the return figure survives because it is the story the team wants to tell. The number that comes out of this process is not a return on investment — it is a return on invoice, dressed up as one.

The reason the collapse happens is not that the team is careless; it is that the measurement brief was never written. The brief that would have stopped the collapse is a single document that names the cost categories the programme owns, the return categories the programme can legitimately claim, the attribution rule that connects the two, and the comparison group that makes the claim falsifiable. Most programmes do not have that document. Most programmes have a budget code and a survey link, and the gap between the two is wide enough to lose an audit.

The rest of this page is a working brief for that document. It sets out the cost categories that survive an audit, the return metrics that survive scrutiny, the attribution rule a programme can defend, and the metrics that look good in a deck and fail when a finance lead reads them back. The list is opinionated rather than exhaustive, because an exhaustive list is the kind of artefact that never gets finished, and a working brief is the kind that does.

Cost categories the brief must capture
Programme labour, programme overhead, production, freight, storage, distribution, replenishment, reporting
Return categories the brief must keep apart
Brand awareness, recipient experience, partner loyalty, sales enablement, internal engagement, retention
Attribution rule
Cohort comparison against an unexposed group, with a defined exposure window
Reporting cadence
Monthly on operations, quarterly on cohort movement, annually on programme economics
What to retire
Impressions on worn merch, social reach on giveaway content, the open-ended NPS line

Four cost categories the brief must capture

The first move in any merch measurement is to widen the cost figure beyond the invoice total. A programme that is measured against invoice cost alone will understate its own spend by between thirty and sixty per cent, because the biggest cost lines are not the lines a supplier charges. The categories below are the ones a finance lead will look for on first review, and the order is the order in which they should be assembled.

Programme labour and overhead

The first category is the cost of running the programme itself, regardless of whether a single garment is produced. Programme labour is the time the marketing or brand lead spends on briefs, supplier calls, artwork coordination and proofs; the time the procurement or sourcing lead spends on quotes, contracts and vendor due diligence; the time the finance lead spends on purchase orders, accruals and supplier reconciliation; the time the warehouse or office manager spends receiving, packing, distributing and reverse-logging the inventory; and the time the in-house designer spends producing artwork, preparing files and chasing approvals. None of this is on the supplier invoice and all of it is on the programme cost. The simplest way to capture it is an internal time log kept for one quarter across all named roles, with each hour marked up at fully loaded cost and totalled into a quarterly programme-labour figure. A programme that does not log this number will report a return-on-invoice figure that is several times the return-on-programme figure, and the gap will become visible the first time the programme is challenged on its scale.

Storage, distribution and reverse logistics

The second category is the holding cost of every unit produced that has not yet shipped to a recipient, plus the cost of moving it to the recipient, plus the cost of bringing it back when it does not. Holding cost is the square-metre charge on a third-party warehouse or the implicit rent on a back-office store room, marked at the prevailing local rate; for an Armenia programme running out of a Yerevan office the implicit rate is the office rent pro-rated by the share of usable storage space, which is rarely counted and rarely zero. Distribution is the courier, the fuel, the route time and the recipient-side handling — for a Yerevan conference run this is straightforward, for a nationwide Armenia rollout or a regional EAEU rollout it becomes a real line. Reverse logistics is the cost of taking units back from recipients who declined them, or from offices that did not use them, or from events that ended with unsold stock; this is the line that is most often forgotten and most often relevant, because a merchandise programme that overshoots its requirement by fifteen per cent is paying twice for the surplus: once to produce it and once to dispose of it.

Production and replenishment economics

The third category is the supplier-side production line and the way it amortises across the programme. Production cost per unit is the figure the supplier quotes — blank, decoration, packaging, freight — set out on the original offer. Replenishment is the additional order placed three months later when the first run is depleted, often at a smaller volume, often at a higher per-unit figure, and often with new setup fees. The total production cost of a programme is therefore the original-order total plus the replenishment-order total, plus any rush or expedite charges that the second run was forced into because the inventory target was set too low. A programme that reports only the original-order invoice is misstating its production cost in a way that compounds year on year, because the replenishment is the part of the cost that grows with the programme maturity. A small programme in year one may show its full production cost on the first invoice; a mature programme in year three will show only sixty or seventy per cent of its annual production cost on the original-order invoices.

Measurement and reporting infrastructure

The fourth category is the cost of producing the measurement itself: the survey platform, the analytics subscription, the time the data or insights team spends on cohort construction, the time the brand team spends writing the report, the time the design team spends producing the visuals that accompany the report, and the time the leadership team spends reading and challenging it. None of this is on a supplier invoice either, and a programme that runs in a small company will absorb it into role time rather than a separate budget line. The reason it belongs on the cost side of the ROI calculation is that the measurement figure should be reported net of measurement cost, not gross — a programme whose measurement cost is high relative to its deployment cost is over-instrumented, and a programme whose measurement cost is low relative to its deployment cost is producing numbers the finance lead will not trust. A useful rule is to cap measurement cost at between five and ten per cent of total programme cost; outside that band the measurement is either too thin to defend or too expensive to justify.

The return side: what counts and what does not

The return categories worth measuring on a B2B programme are not the categories a consumer campaign would measure. A B2B programme is judged by what it does to a defined audience — employees, channel partners, key accounts, conference attendees, board-level relationships — and the return categories that matter are the ones that have a measurable effect on that audience. Six return categories survive scrutiny; everything else should be retired or moved to the qualitative side of the report.

Brand awareness is the first and is the easiest to overstate. A piece of merch worn in public is exposure, and exposure has a value, but exposure has no measurable return unless it carries a trackable element — a tagged landing page, a QR-coded item, a unique referral identifier, a recipient-only domain — that converts the exposure into traffic the brand can attribute. Without that trackable element, the exposure number is vanity and should be reported as such. Recipient experience is the second and is the most direct. Surveys sent within seven days of receipt, scored on a five-point scale with a defined response threshold, produce a number that the brand team can defend and the finance lead can recognise as a customer-satisfaction metric.

Partner loyalty is the third and is where a B2B programme differs from a consumer one. A channel partner who receives a kit at the start of a quarter is more likely to participate in the next quarter than a comparable partner who does not. That claim is testable with a cohort comparison and should be, because the cost of a partner churn event is high enough to justify the measurement. Sales enablement is the fourth and is the most expensive return category to measure, because it requires a comparison between accounts that received a sales kit and accounts that did not, and the comparison is confounded by deal size, segment, account maturity and seasonality. Internal engagement is the fifth and is measured against an internal control group — an office that received a kit against an office that did not — for retention, eNPS and absenteeism signals. Retention is the sixth and is the slowest-moving of the six, measured on a twelve-month horizon with a cohort that was exposed at month zero.

The six return categories are not equally easy to measure. Brand awareness is easy in the wrong direction — exposure is countable and valueless — and hard in the right direction. Recipient experience is easy and produces a defensible number. Partner loyalty is testable but requires a control. Sales enablement is hard and confounded. Internal engagement is testable on a small group. Retention is testable but slow. A programme that wishes to claim a return on investment should measure the categories it can defend and report the rest as qualitative observation, labelled as such, instead of inflating them into numbers.

Attribution, incrementality and the cohort question

The attribution rule on a B2B merch programme is the move that separates a defensible ROI from a memorable anecdote, and it is the move most programmes skip. The reason programmes skip it is that a rigorous attribution requires an unexposed control group, and most programmes are run on populations that are not designed to leave a control group behind. A conference run is a population that arrives together; a partner programme is a population that is, by definition, every active partner; an internal engagement programme is the entire staff. Without a control group the only honest claim a programme can make is descriptive — what the exposed group did — and the inference from descriptive to causal has to be reported with the caveats the description deserves.

A cohort comparison is the discipline that changes this. The exposed cohort is the group that received the merch; the control cohort is the matched group that did not; the comparison period is the window between receipt and the next observable behaviour. The two cohorts should be matched on as many observable variables as possible — for a partner programme, on prior spend, prior tenure, tier and segment; for an internal programme, on role, level, tenure and office; for a conference programme, on attendee role, company size and the duration of attendance. Where a match is impossible, the cohort design has to be reported as quasi-experimental rather than experimental, and the claim that comes out of it should be calibrated accordingly.

Incrementality is the second concept and is the one most often confused with attribution. Attribution asks which observed change was caused by the merch. Incrementality asks whether the change would have happened anyway, in the absence of the merch. The two questions look like the same question and they are not. A partner who places a second-quarter order after receiving a kit has been attributed to the kit; the question is whether they would have placed it anyway, and the answer is sometimes yes, sometimes no, and only a control cohort reveals the difference. A programme that reports attribution without incrementality is reporting the wrong number, and a programme that reports both is reporting a number finance can sign off on.

The honest caveat on any merch attribution is that merch is a low-frequency, long-tail channel in most B2B mixes. The presence of a kit in a partner mailbox is one of perhaps a hundred inputs in the next purchase decision, and isolating the kit contribution requires a measurement brief that explicitly rules out the other ninety-nine. Most programmes do not have that brief. The programmes that do are the ones whose ROI figures survive a finance review.

Six metrics that survive a finance review

The metrics below are the ones a B2B merch programme should report when the brief is mature. Each is observable on the programme itself, defensible against a control, and meaningful at the scale of the programme. None of them is novel, all of them are reachable, and each has been used on at least one B2B programme we know of. A programme that reports against all six has a defensible dashboard; a programme that reports against three has a defensible starting point; a programme that reports against one is reporting a single signal and should label it as such.

Recipient experience score. A five-point scale, sent within seven days of receipt, on a defined sample with a defined response rate, scored against a benchmark from the prior period. The benchmark is what makes the number defensible; a programme that reports its score without a benchmark is reporting an absolute number against a moving target. Cohort retention delta. The percentage of the exposed cohort that is retained at twelve months minus the percentage of the control cohort that is retained at twelve months. Internal for an employee programme, contractual for a partner programme. The delta is the number, not the absolute retention, because the absolute retention is dominated by forces outside the programme.

Partner reorder rate. For a channel partner programme, the number of partners in the exposed cohort who place a subsequent order within the next two quarters divided by the number who did not, against the matched control. Conference lead-to-meeting conversion. For a conference programme, the percentage of recipients who book a follow-up meeting within thirty days divided by the matched control, scored on the lead-to-meeting conversion rather than on the lead-to-revenue conversion because the latter is too confounded by deal-cycle length. Internal engagement signal. An eNPS-style or participation signal on the exposed office against the unexposed office, measured at the next survey interval and reported with a defined comparison window.

Programme cost per retained unit. The total programme cost divided by the number of retained units at twelve months, where a retained unit is a recipient who is still in the relationship the merch was sent to support. The metric is useful because it puts a cost-per-retained-unit figure into the budget conversation in a language finance recognises, and because it forces the cost side and the return side onto the same axis. A programme whose cost-per-retained-unit falls year on year is improving on its own terms; a programme whose number rises is producing more spend per retained recipient and should be investigated before the next budget round.

Five vanity metrics to retire from the dashboard

The metrics on the second list are the ones a B2B merch programme often reports because they are easy to capture, not because they are defensible. They are the items that look good in a slide and fail when a finance lead asks how they were attributed. Retiring them is not a loss — it is a redirect of measurement effort onto metrics that survive scrutiny.

Impressions on worn merchandise. A garment worn in public produces an exposure number, but the number is unfalsifiable, untagged, and unaffiliated with any measurable behaviour. Report it as descriptive, not as a return signal. Social reach on giveaway content. Reach is what the platform counts when an item is photographed; it is not a return signal for the programme. Use it for the creative review, not for the ROI review. Open-ended NPS lines. “How do you feel about the kit?” produces a quote, not a score. A quote belongs in the qualitative section of the report; a number belongs in the quantitative section; they should not be mixed.

Pre-existing loyalty gains attributed to the kit. A partner who places a second order after receiving a kit may have placed the order anyway. Without a control cohort the attribution is the wrong inference. Conference dwell time at a booth. Dwell time at a merch giveaway is a measure of booth staff competence, not of merch programme effectiveness. A better question is the lead-to-meeting conversion on the recipients who received the merch, not the dwell time at the booth where it was given away.

The list is not exhaustive. The general rule is simple: if the metric would still be reported under a deliberately falsified programme — that is, if a programme designed to look successful but not actually working would still produce the number — the metric is vanity and should be retired. Vanity metrics are not dishonest, they are simply decoupled from the thing the programme is meant to do, and a dashboard full of them is the kind of artefact that loses budget the next time the budget is contested.

Reporting cadence finance and brand can both live with

The reporting cadence on a B2B merch programme should be set by the cadence of the decisions the report feeds. Operations data — production status, distribution status, inventory levels — should be reported at the frequency the operations team can act on, which for most programmes is weekly during a rollout and monthly outside it. Brand and cohort data — recipient experience, retention delta, reorder rate — should be reported at the cadence the brand team can interpret, which for most programmes is quarterly. Programme economics — total cost per cohort, cost per retained unit — should be reported annually with a mid-year informal update.

The cadence discipline matters because the wrong cadence produces the wrong behaviour. A monthly cohort report forces the team to interpret movement that is not yet interpretable and produces false signals. A quarterly operations report produces a programme manager who cannot see the production issue in time to fix it. The answer is to split the cadence by the kind of data, rather than running the whole report on the most convenient frequency, and to decide in the brief which data goes to which audience.

Finance is the audience that is most often forgotten in the cadence conversation, and the one that is most often blocked by a report that lands out of cycle. Finance reviews the budget against actuals on a calendar that is set six months in advance, and a programme that wants its number to be trusted should land its annual figure inside the same calendar. That constraint alone shapes the brief: when the programme starts in March, the annual figure has to be ready by January, which means the measurement infrastructure has to be in place by September, which means the brief has to be written in June. Cadence is not an afterthought; it is the calendar the brief has to fit.

Writing the measurement brief the programme deserves

The measurement brief is a one-page document, owned by the programme lead, signed by finance, brand and procurement. The brief has six sections, each a paragraph long, and the totality is the document the next budget conversation will quote. The sections are: scope — the population the programme covers and the population the control covers; cost categories — the eight to twelve lines the cost figure will be assembled from; return categories — the four to six defensible signal lines; attribution rule — the cohort design and the comparison window; reporting cadence — the calendar by data type; and retirement list — the metrics the programme will no longer report. None of the sections is long, all are essential, and the brief should be updated annually on the same date the budget is set so that the document and the number move together.

A useful test of the brief is the false-programme test. Imagine the programme is producing all the right data but no real return, only the appearance of one. Would the report still come out positive? If yes, the brief has at least one vanity metric on it, and the vanity metric should be retired before the next measurement cycle. A brief that fails the false-programme test produces a number that finance will challenge; a brief that passes it produces a number finance can sign.

Questions buyers ask about merch measurement

What is a defensible benchmark for the recipient experience score on a B2B merch programme?

A defensible benchmark is the prior-period score of the same programme, scored on the same five-point scale, sent at the same interval, against a comparable sample. Cross-programme benchmarks are unreliable because the populations and contexts differ, and an industry-wide benchmark is usually an artefact of the survey vendor rather than of the category. The right discipline is to report the prior-period figure alongside the current-period figure, with the absolute number and the delta. A programme whose score moves up by a defined amount on the same scale is improving; a programme whose absolute number is high but unmoving is plateauing.

Is it acceptable to report a merch programme ROI without a control cohort?

It is acceptable to report descriptive numbers — what the exposed cohort did — without a control, and to label the report as descriptive. It is not acceptable to report a return-on-investment number without a control cohort and to label that number as ROI, because ROI is a counterfactual claim and the control is what makes it falsifiable. The cost side of the equation does not need a control; the return side does. A report that runs without a control should describe and should not conclude, and the brief should say so before the first measurement cycle.

What should replace impressions on worn merch if the metric is retired?

A tagged-exposure metric: a QR-coded item that points to a tracked landing page, a unique recipient code that registers at a portal, or a domain name that is unique to the programme. Each converts an exposure into a trackable signal the brand can attribute. The signal will be a fraction of the vanity figure, but the fraction will be honest and the rest of the vanity figure can retire. Programmes that move from impressions to tagged exposure tend to lose the headline number and gain the audit defensibility, which is the right trade.

Is Corp-Merch a partner or official supplier of the brands, frameworks, surveys, sample sizes or methodologies referenced on this page?

No. Corp-Merch is an independent B2B merchandise supplier operating in Armenia and the wider EAEU. References to attribution methods, cohort designs, sample-size rules, scoring instruments, survey platforms, brand-tracked mechanisms or industry benchmarks are informational only and imply no partnership, sponsorship, endorsement or official-supplier relationship with any of them. Nothing on this page constitutes financial, marketing or research advice; every programme should be measured on its own brief, and every brief should be signed by the finance, brand and procurement leads who own the number.

This page is written for marketing, brand, procurement and finance leads running a custom B2B merchandise programme in Armenia and the wider EAEU. Corp-Merch is an independent B2B merchandise supplier. The measurement categories, attribution rules, sample-size references, scoring instruments and reporting cadences described here are general practice for this category as we understand it and are not financial, marketing, statistical or research advice; each programme applies its own brief, its own sample design and its own internal review. No brand, framework, survey instrument, analytics platform, sample-size rule or attribution methodology referenced here is affiliated with, endorsed by, or an official supplier to Corp-Merch, and any figure or range on this page should be treated as a planning estimate until it is confirmed in a written offer, a signed brief and an approved measurement calendar for the specific facts of the programme.