The honest measurement glossary

By the HonestTag team · Published July 12, 2026

Ad measurement runs on words that sound precise and are not. "Conversion" means whatever the platform counting it wants it to mean, and "CAC" changes value depending on who is in the denominator. This page pins down what each term means inside HonestTag, so when a report says NCAC or a verdict says Hold, you know exactly what was computed and from what.

Core metrics

A note on spelling. HonestTag writes these metrics NCAC and NMER everywhere: the app, this site and the docs. You will also see them written nCAC and nMER, here and elsewhere in the industry, where the small n stands for new customer. They are the same two numbers.

MER

Marketing efficiency ratio: total net revenue divided by total ad spend. Net revenue means ex-tax, net of discounts and returns, plus shipping collected. MER is the slow company scoreboard. It moves with everything, so it tells you how the whole business is doing, not which channel did it. Full guide: What is MER?

NMER (aMER)

New-customer net revenue divided by total ad spend. Also called Acquisition MER: one number, two labels. This is the prospecting metric: it only rises when ads create customers you did not already have. Full guide: What is NMER?

NCAC

New-customer acquisition cost: total ad spend divided by new customers acquired. Returning buyers do not belong in the denominator, which is why NCAC usually runs higher than blended CAC, and reads truer. Full guide: What is NCAC?

First-Click NCAC

HonestTag shows two first-click NCAC figures. On the scorecard, First-Click NCAC is total paid spend divided by all new customers whose first recorded ad click was a paid platform. On the Gap page, the per-channel figure is that channel's spend divided by the new customers whose first recorded ad click came from it; campaign verdicts use the same calculation at campaign level. The channel and campaign figures are what HonestTag verdicts judge against your ceiling. Guides: NCAC and first-click attribution.

AOV

Average order value: net revenue divided by all orders, ex-tax. HonestTag computes this blended AOV and uses it in the suggested NCAC ceiling when no explicit target is set. First-order AOV is a separate concept a merchant may compute from new customers' first orders. Full guide: What is AOV?

NCAC ceiling

The most you can pay for a new customer. It is your explicit target if you set one; when you have not, HonestTag derives it as AOV x gross margin percent x an LTV multiple. Every verdict cites which source it used, so the bar you are judged against is never a mystery. Guides: NCAC and verdicts.

Attribution

First-party tracking

Data collection on the store's own domain, under the store's own consent settings. The click and the order are recorded by the store, for the store, not by a third party grading its own ads. Full guide: first-party tracking.

First-click attribution

Credit goes to the ad click that first brought the customer. It answers the acquisition question, what created this customer, instead of rewarding whatever touch happened to sit nearest checkout. Full guide: first-click attribution.

Attribution model

The rule deciding which touch gets credit for an order. HonestTag reports exactly one model, first-click new-customer, sends every verified purchase whatever the model says about it, and shows read-only comparisons to last-click and any-click so you can see how much the choice of rule moves the numbers. Full guide: attribution models.

Last-click

Credit goes to the touch nearest checkout. It systematically over-credits retargeting and branded search, because those touches cluster around orders that were often already decided. HonestTag shows it as a comparison, not as truth.

Any-click (participation)

An order counts once for every platform that touched the journey. The columns overlap by design and do not sum to your order count. The view answers "was this platform present," not "what caused this order."

Evidence record

The stored proof behind every attribution: click id, timestamps, match method, new/returning classification, and any refunds and adjustments. The pricing page names the same thing the proof record. A number that cannot show its record is an estimate wearing a suit.

Click-to-order lag

The days between a new customer's first click and their first order, bucketed per channel: same day, 1-3, 4-7, 8-14, 15-30, and 31+. It tells you how long a channel needs before its results are readable. Full guide: click-to-order lag.

Verdicts and evidence

Verdict

The weekly Scale, Hold, or Cut read per channel, and per campaign on Truth and up. It is computed by a fixed, published rule from your own numbers, issued Mondays, immutable for the week, and always shown with the numbers that produced it. Full guide: verdicts, or see them in the product.

No verdict

The fourth state. When an evidence floor is not met, HonestTag says so and names the floor and what unlocks it, instead of inventing a verdict from thin data. Silence with a reason beats confidence without one.

Evidence floor

The published minimums that gate every verdict: store and channel new-customer counts, spend, campaign age, data freshness, and a resolvable NCAC ceiling. Below the floor, the answer is No verdict, never a guess dressed up as one.

New vs returning purchase signals

Google Ads receives the standard purchase conversion plus HonestTag's classified actions (new customer, returning customer, first-click new customer). Meta receives the classified custom events, while its standard Purchase is sent when the merchant answers that Shopify's Facebook & Instagram channel is off, set to Standard, or not installed. They exist so campaign optimization can learn from real acquisition only, rather than from repeat buyers it merely intercepted. Full guide: new vs repeat signals.

Reports and tools

The Gap

Platform-claimed conversions next to the store's first-click new-customer truth, channel by channel. The distance between the two columns is what self-graded reporting has been costing you.

Campaign matching

Joining platform campaigns to your orders through utm_campaign: id match first, then normalized name, else unmatched. Unmatched spend and unmatched orders always render, and an account-level adjustments row makes the table add up to the channel topline. Nothing quietly disappears. Full guide: campaign matching.

Cohort

Customers grouped by first-order month. Cohort payback is the month cumulative revenue per customer crosses acquisition cost, the point where a cohort has paid for itself. Full guide: cohort payback.

LTV by channel

Cohort revenue curves split by the acquiring first-click platform, stamped at first order, with a 30-customer floor per cell so a thin cell never masquerades as a trend. Full guide: cohort payback.

LTV projection

A young cohort's future months, projected by chaining the median growth ratios of the store's own older cohorts: 3+ donor cohorts of 30+ customers per step, capped at month 12, and never extended beyond the history the store has actually observed. Projections render separately and are never blended into actuals. Full guide: LTV projection.

Tracking link

A short link on the store's own domain that redirects to a store page with the link's UTMs applied. It uses a 302, so clicks keep counting and edits keep working. Every click is counted and lands in the same first-party attribution and order proof as an ad click. Built for influencer, organic, and email-signature placements ads cannot tag; every paid plan includes some, and caps rise by plan. See tracking links in the product.

Email touch

An email or SMS click, ingested from the merchant's own connected email platform. This is coming next, not live yet. When it ships, it will appear in the order proof as an assist and in an Email channel row, but email will never take first-click credit. HonestTag reads click-engagement events only, never lists, segments, or profile properties. See email touches in the product.

Refund retraction

When an order is refunded, Google Ads and Microsoft Advertising receive the available retraction or restatement, and Klaviyo and Google Analytics 4 receive a refund event. HonestTag does not send Meta, TikTok, Pinterest, Snapchat, Reddit or OpenAI Ads a refund or retraction event; the refund is recorded in order proof. Store-level revenue metrics are reduced by the refund.