Armenia · cost per impression · denominator · sightlines · AMD numerator · finance-grade reporting

Cost Per Impression for Merch in the Armenian Market

The cost-per-impression figure is imported into Armenia from media planning, where impressions are counted by servers and priced in a single currency. Neither of those conditions holds for merchandise bought here. Counting happens in showrooms, at conference stands, in branch offices and at kitchen tables in marzes; a meaningful share of the spend is budgeted in dram while the production and freight lines arrive in another currency entirely. This page works through the denominators that survive those conditions, the four places the figure most often collapses in an Armenian programme, and the reporting shape that lets a marketing lead put the number in front of finance without it being dismantled on the spot.

Talk to us about measurementAll B2B notes

Why the imported version of the metric does not survive contact with a local programme

Two assumptions sit underneath the imported version of this metric, and both fail here. The first is that impressions are countable. In paid media, an impression is a server event: it happens, it is logged, it is billable, and the count is produced by infrastructure that has no opinion about the buyer’s product. For merchandise in Armenia, the equivalent count depends on a person being in a room with the item at the same time as a person who would notice it. At a Yerevan trade stand that is a conversation-length encounter between two people and an item held in a third person’s hands. In a branch office in Gyumri or Vanadzor it may be a rack on a wall for eleven months. In a courier-to-home drop it may be a box opened in a hallway and discarded the same evening. None of those produces a log, and the team that reports an impression count without saying which of them it counted has not measured anything — it has estimated, which is a legitimate thing to do provided the label says so.

The second assumption is that the cost and the benefit are denominated in the same currency and visible together. In Armenia the budget is usually set in dram and approved in dram, while blanks, decoration and often freight are priced in a foreign currency and converted at a rate that moves between the moment of quotation and the moment of payment. A programme quoted in March and invoiced in June has two different dram figures for the same goods, and neither of them is wrong. What breaks the metric is that a cost-per-impression figure presented in dram inherits whichever conversion was used, without saying so, and a reader comparing it to the previous quarter’s figure is comparing two conversion rates as well as two programmes. The numerator therefore needs a stated rate and a stated date, not just a number.

There is a third, quieter problem: the audience that sees merchandise in Armenia is frequently not on the buyer’s list at all. A giveaway handed to a visitor at a stand is seen by the visitor, by the person they speak to afterwards, and by whoever is standing nearby; a gift set taken home by a partner’s employee is seen by a household; an item left in a car is seen by nobody and used by nobody. Counting only the named recipient understates the reach by a wide margin, while counting every person who could theoretically see the item inflates it into something indefensible. The usable answer sits between the two, and finding it is a local exercise rather than a formula — which is why the denominator section below is split by setting rather than by arithmetic.

Counting method
Observation or estimate, never server-logged; the label must say which
Currency split
Budget in AMD; production and freight often priced elsewhere
Conversion disclosure
Rate and date stated alongside the numerator
Unlisted audience
Household, queue, stand visitor and colleague are outside the recipient list
Setting split
Stand, office, courier-to-home and vehicle produce different denominators
Working rule
Report a range with a named low and high, not a single point estimate

Four denominators that hold up in an Armenian programme

A defensible cost-per-impression figure in this market is usually built on four denominators rather than one, because four distinct settings produce four distinct kinds of viewing. Rather than average them into a single number, the better practice is to compute the figure once per setting and report the range, letting the reader see which environment is carrying the programme. Each denominator below is described with what it counts, what it deliberately excludes, and the mistake it is designed to avoid.

The recipient denominator

This is the smallest and the most defensible: it counts only the named people who received an item, one viewing each, on the basis that every recipient is at minimum a person who has the object in their possession. It is the denominator to use when the item is a personal-use object — a bottle, a notebook, an apparel item intended to be worn — because possession is close to the truth and nothing has to be assumed. Its weakness is that it produces a high cost per impression, which makes the programme look expensive next to a stand giveaway. That is not a flaw in the measure; it is the measure doing its job. A recipient-only figure is the floor of the range and the one that will still be defensible in two years when the recipient has become a sceptic of the programme.

The household and workplace denominator

An item that lives in a home or a shared office is seen repeatedly by people who are not recipients, and in this market that multiplier is larger than in a single-office market because staff who work on site travel between offices and take items with them. The useful convention is to count a modest, explicitly stated secondary audience — a household for a home-delivered gift, a small team for an office-distributed item — and to apply it consistently across programmes. The consistency matters more than the exact number: a figure built on a stated secondary multiplier can be adjusted by a reader who disagrees with it, whereas a figure built on an impressionistic estimate cannot. The failure this denominator prevents is the programme that quietly reports only recipients and is then compared with a programme that quietly reported recipients plus staff.

The event-day denominator

For anything distributed at a stand, at a conference, or at a Yerevan trade exhibition, the denominator is time, not people. An item held at a stand for a three-day show with steady footfall accumulates far more viewings than a stack of the same item on a desk accumulates in a month, and the difference is large enough to make a programme either look excellent or worthless depending on which environment it is measured in. The workable approach is to define a minimum engagement — a stop, a handled item, a completed conversation — and to state that as the counted unit rather than counting passers-by. Event-day programmes should also be reported against a separate cost numerator, because stand staffing, travel, and on-site collateral are costs that never appear on a merchandise invoice and that make an apparent cost-per-impression advantage evaporate once they are added back.

The carry-over denominator

Some items in an Armenian programme are not consumed on receipt and are not stored either — they are worn, carried, or left in a vehicle, which makes their life unpredictable and long. The carry-over denominator handles these by splitting the item into two populations: a primary population that is retained and used, and a secondary population that is retained but barely used. Reporting the two separately is usually more informative than averaging them, because it tells the programme whether the specification is wrong. An item with a large barely-used population is an item whose design or category was not matched to how the audience actually lives, and that is a finding worth more than a tidier single figure. The mistake this denominator prevents is the long-lived average, which smooths a real weakness into a comfortable number.

Four places the figure collapses in an Armenian programme

The metric fails in predictable ways, and knowing them is faster than discovering them. The four below account for most of the bad numbers that circulate in marketing decks in this market, and each has a specific corrective move.

The giveaway counted as a keep. A giveaway handed out at a stand is consumed by the event and usually leaves the building in a bag that is never opened again. Reporting it on the same denominator as an item that is kept for a year is the most common distortion in the category, and it is the distortion that flatters the programme most, because a large giveaway volume produces a very small cost per impression. The corrective move is to put giveaways in their own line, with their own denominator, and to report them as a separate objective. A stand that produces high volume and low carry-over is a lead-generation activity; describing it as a brand-impression activity sets up a measurement argument that the numbers cannot win.

The currency drift. A figure quoted in dram at quotation and in a converted currency at production, then compared across quarters, is comparing exchange rates as much as programmes. The corrective move is to fix one denomination for the series, state the rate and the date it was taken, and hold that rate constant for comparison purposes even when invoicing moves. A marketing lead who reports a single-quarter figure in a floating denomination is reporting a currency movement and calling it a merchandising result.

The unlisted-audience omission. When a programme distributes to a named list but the items are then shown in offices, at events, and at home, the reported denominator understates reach by a factor the programme cannot defend without saying so. The corrective move is to apply one named secondary convention across the whole series, and to state in the same sentence that the figure is a floor rather than a ceiling. Understating on purpose and labelling it is defensible; understating silently is not.

The single-year comparison. Comparing an Armenian programme against a paid-media benchmark from a different market, a different year, or a different currency is the most tempting and least defensible move, because the benchmark is a number the team did not produce and cannot audit. The corrective move is to build an internal series first — the same programme, the same settings, the same denominators, quarter after quarter — and to use any external benchmark only as a directional note, never as the justification for a budget line.

Building a figure that survives a budget meeting

The purpose of the metric is not to prove that merchandise is cheap. It is to give a marketing lead a defensible sentence when the budget line is questioned, and that requires a shape rather than a number. A figure presented alone invites three questions, and a figure presented with its structure invites none of them. The structure below is what to put in the document.

Present four numbers, not one: the landed cost numerator with its conversion date, the recipient count, the named secondary multiplier, and the resulting cost-per-impression range across the four settings. Then attach one paragraph naming the counting method in a single sentence — observation, survey, or stated assumption — because a reader who does not know which one was used cannot evaluate any of the four numbers. Add the comparator on the same page: the same programme in a previous period, or a sibling programme with the same denominators. A figure with a comparator is a measurement; a figure without one is an assertion.

Then state the exclusions plainly, in the same document rather than in a footnote. Stand staffing, travel, on-site collateral, internal coordination time, and the amortised share of any setup charge are all real costs and all easy to leave out of a per-unit line. A cost-per-impression figure that excludes on-site costs for a stand programme is not a cheaper programme; it is a programme where the costs went somewhere other than the metric, and finance will find them. Writing the exclusions down is what makes the figure usable in a budget argument rather than merely internally satisfying.

Finally, decide in advance what result would change the programme. A metric with no decision attached to it is a reporting exercise, and reporting exercises get cut the first time the budget is tight. Naming the threshold in advance — the cost-per-impression level above which the specification changes, the retention reading at which the item is reconsidered, the share of giveaways above which the objective is restated as lead generation — is what turns the number into a control. Those thresholds should be set before the figure arrives, not after, and the honest version of the practice is to record them and keep to them even when the result is inconvenient.

Questions buyers ask about cost per impression in Armenia

Is there a way to measure merch impressions that does not rely on assumptions?

Not at the level of a single item, and pretending otherwise is how most figures in this category go wrong. What can be counted reliably is narrower: how many items were issued, how many were still held at a follow-up point, how many unprompted requests came back, and how many times a stand item was handled rather than merely passed. Those four are auditable and repeatable. The multiplier applied to reach beyond the recipient is always an assumption, and the honest treatment is to name the assumption, hold it constant across the series, and show a reader how sensitive the final figure is to it. A metric whose answer moves sharply when a reasonable assumption moves is a metric that should be reported as a range with the sensitivity stated.

How should giveaways and retained gifts be reported if they sit in the same budget line?

Separately, with separate denominators and separate objectives, even when they share a budget line. A giveaway is judged on reach and on what it produced for the business that followed; a retained gift is judged on whether it was kept and used. Merging the two produces a number that is high on volume, low on cost, and meaningless, because the two populations behave so differently that the average describes neither. The practical reporting shape is one table with two rows, so that the budget is visibly accounted for while the objectives stay distinct, and so that a reviewer can see immediately which half of the line is doing which job.

Our budget is approved in dram and production is priced in another currency. How do we handle that?

Fix one denomination for the measurement series, state the rate and the date it was taken, and hold it constant for comparison even where invoicing moves. This keeps quarter-on-quarter comparison meaningful and stops an exchange-rate movement from being presented as a merchandising outcome. Where the exposure is large enough to matter, the more useful control is at the procurement end: agreeing the conversion basis and the point at which it is fixed as part of the order, rather than discovering it at invoice. That is a commercial point rather than a measurement one, but it determines whether the cost numerator in the metric is a number or a range.

Is Corp-Merch the official supplier or affiliate of any measurement body, media convention, or industry reference on this page?

No. Corp-Merch is an independent B2B merchandise supplier operating in Armenia. The denominators, counting conventions and reporting shapes described here are a reasoned working approach to a metric that this category has borrowed from media planning; they are not a recognised standard, an audit method or a certification, and they imply no partnership, sponsorship, endorsement or official-supplier relationship with any body. Nothing on this page guarantees a measurement outcome on a future programme.

This page is written for marketing, brand and finance leads in Armenia who need a cost-per-impression figure they can defend. Corp-Merch is an independent B2B merchandise supplier. The four denominators, the counting conventions and the reporting shape set out here are a reasoned synthesis of how the metric behaves in this market as we understand it; they are not marketing research, accounting advice or a measurement standard, and any figure produced with them should be calibrated to the actual observation and survey discipline available on the programme. Any reference to a named convention, method or published category is informational only and implies no partnership, endorsement or official-supplier relationship with the organisation that publishes it.