E-Commerce Ad Intelligence Glossary: 50 Terms Every Operator Must Know

Why publish a glossary

Operators reading this glossary fall into two camps. The first is new to ad intelligence — they have heard the terms but want a canonical short list with operator-side definitions, not vendor marketing copy. The second is experienced but tired of explaining the same terms to junior team members; they want a reference URL to send.

This glossary is written for both. Each term gets a definition of two to four sentences. Operator-side framing — what the term means when you are deciding what to do — wins over textbook definitions. Where a term is CommonWealth Ops specific (trendingscore, viewsband, Whisper ASR), we say so explicitly. Where the platforms (Meta, TikTok) define the term, we cite the public surface.

The list is intentionally fifty terms. Longer lists become unreadable; shorter lists skip too much. A future quarterly update may rotate two or three terms as the field evolves.

A

Ad fatigue Loss of creative effectiveness from overexposure to the same audience. Symptoms: CTR drops, hook rate drops, frequency rises, CPM rises as the platform de-prioritises a tired ad. Fatigue typically appears between days 14 and 45 of an ad's life depending on audience size and budget. Operators rotate hooks (not entire ads) to extend the run.

Ad set The Meta organisational layer between Campaign (objective) and Ad (creative). An ad set defines audience, placement, budget, schedule, and bid strategy. TikTok calls this an Ad Group. Ad sets are the unit at which most operator-level decisions land — pause, scale, duplicate.

Advertiser identity The brand name and verified page shown on the public ad surface. Operators reading the ad library to identify competitors track the advertiser identity, but the identity can be a sub-brand, a regional variant, or a media buyer's holding entity — the surface name is not always the parent company.

AOV (Average Order Value) Total revenue divided by total orders over a defined window. A higher AOV unlocks higher CAC tolerance — if AOV is EUR 80 instead of EUR 40 the same product can sustain double the per-acquisition cost. Bundle offers, free-shipping thresholds, and post-purchase upsells are the three usual levers.

Attribution model The rule deciding which marketing touchpoint gets credit for a conversion. Last-click, first-click, linear, time-decay, and data-driven are the common variants. Different models attribute the same revenue to different channels — comparing campaign performance across models can swing decisions by 30+ percent.

B

B2B advertising Advertising to businesses rather than consumers. B2B has longer sales cycles, smaller addressable audiences, higher per-deal value, and stronger reliance on LinkedIn and trade publications. Most ad-intelligence tools (CommonWealth Ops included) are B2C-first; we cover B2B distinctions in our B2B competitive intelligence post.

Brand recall The percentage of an exposed audience that can identify the brand after exposure. Measured by survey or post-ad lift study. High brand recall is the load-bearing top-of-funnel metric for awareness campaigns; performance-only operators often ignore it and lose long-term LTV.

C

Cohort analysis Grouping customers by their acquisition period (week, month, quarter) and measuring how each cohort behaves over time — repeat purchase rate, LTV at 90 days, churn. Cohort views catch what aggregate dashboards hide: a single bad quarter of acquisitions can drag aggregate LTV down for two years.

CPM (Cost per Mille) Cost per 1,000 impressions. The auction-side input to most ad pricing. CPM rises with audience competition, time of year (Q4 doubles in most markets), and creative quality (low quality CPM rises as the platform deprioritises). CPM alone says nothing about conversion — pair with CTR and CVR.

Creative testing The structured practice of running multiple creative variants against the same audience to identify winners. Best practice in 2026: test hook variants (first 3 seconds) on a single base creative before testing entire creative variants. Hook tests are cheaper than full creative tests and reveal more signal per dollar.

CTR (Click-Through Rate) Clicks divided by impressions. A back-end metric reported inside the advertiser account, not visible on the public ad library. Industry-median CTR for direct-response Meta video ads in 2025 was around 0.9-1.5 percent; TikTok median was 1.0-1.8 percent. Above-median CTR is usually the hook combined with the offer, not the targeting.

Custom audience A list of known users — site visitors, email subscribers, past purchasers, app users — uploaded to Meta or TikTok for targeting. Custom audiences are the seed for retargeting AND for lookalike audiences. Privacy-policy compliance (consent, retention) is non-trivial — many operators under-document this.

D

Direct response Advertising designed to drive immediate conversion (purchase, signup, install) rather than long-term brand recall. Direct response ads carry explicit CTAs, scarcity or urgency triggers, and measurable conversion windows. CommonWealth Ops's capture is heavily direct-response weighted because the public libraries surface more direct-response ads than brand ads.

DTC (Direct-to-Consumer) Selling direct to the end consumer without a retailer or wholesaler intermediary. DTC brands own the customer relationship, the pricing, the brand voice, and the data — at the cost of running their own acquisition. Most of CommonWealth Ops's subscriber base is DTC e-commerce in fitness, skincare, supplements, and adjacent niches.

E

Excluded audience An audience explicitly defined to NOT receive ads — past purchasers (for acquisition campaigns), staff and team members, competitors. Exclusions are the most often-misconfigured Meta setting; an unintentional exclusion can wipe 30 percent of qualified reach.

F

Frequency The average number of times a single user has seen an ad over the campaign window. Frequency rising above 4-6 for direct response often signals ad fatigue. Brand campaigns sustain higher frequency by design. Frequency is reported in the advertiser account; the public libraries do not surface it.

Funnel The metaphor for buyer progression through awareness → consideration → conversion → retention. Each stage has different ad creative requirements, different conversion rates, and different cost-per-action expectations. Operators conflating funnel stages usually under-invest in the top of the funnel and run out of audience at the bottom.

Funnel stage targeting (TOFU/MOFU/BOFU) Top of funnel (TOFU) targets cold audiences with awareness ads; middle (MOFU) targets warmed audiences with consideration ads; bottom (BOFU) targets retargeting and lookalike audiences with direct-response ads. The split of budget across stages is the load-bearing operator decision; the typical mistake is over-investing in BOFU and starving TOFU until the pipeline empties.

H

Hook The first one to three seconds of an ad — the moment that decides whether the user keeps watching or scrolls past. The hook is the single highest-leverage element of any video ad. CommonWealth Ops's archetype classification (identity, result, problem, social proof) is hook-level not whole-ad-level.

Hook rate The percentage of viewers who watch more than three seconds of a video ad. Hook rate is a front-end metric — measurable on the public ad surface or via the advertiser's video-retention chart. Strong hook rate is necessary but not sufficient for strong CTR; weak hook rate is sufficient for weak CTR.

Hook rotation The practice of swapping the first 1-3 seconds of an ad while keeping the body and CTA constant. Hook rotation extends an ad's life by resetting fatigue at the highest-impact element. Operators rotating only entire creatives miss the cheaper, higher-signal variant.

I

Impression A single ad view, regardless of engagement. One user can produce many impressions of the same ad over a campaign. Public libraries (Meta, TikTok) do not disclose impression counts to outside operators. CommonWealth Ops uses run duration and creative variants as observable proxies, not impressions.

L

Landing page The destination URL the ad clicks through to. Landing-page quality (load speed, message match, conversion design) is responsible for as much variance in CVR as the ad itself. Operators auditing competitor landing pages benefit from tools that capture the post-click experience — CommonWealth Ops captures the ad surface only; competitor landing pages are out of scope.

Lookalike audience An audience generated by Meta or TikTok to share characteristics with a seed audience (typically purchasers). Lookalikes are the top-of-funnel acquisition engine for most DTC brands. Lookalike quality depends on seed audience size and seed audience purity — a 10-user seed produces a noisy lookalike; a 5,000-user seed produces a meaningful one.

LTV (Lifetime Value) Total net revenue from one customer over their entire relationship with the brand. LTV-vs-CAC ratio (target 3:1 or better) is the financial guardrail for ad budgets. Subscription products and high-repeat consumables have higher LTV than one-off purchases, sustaining higher CAC.

M

Meta Ad Library The public archive of every ad currently or recently running across Meta's platforms (Facebook, Instagram, Messenger, Audience Network). The library is searchable by advertiser, country, and ad text. It is the structural source for most ad-intelligence tools. We document what the library exposes — and what it does not — in our Meta Ad Library API post.

MER (Media Efficiency Ratio) Total revenue divided by total ad spend across all paid channels. MER is the operator-side aggregate health metric — more meaningful than per-campaign ROAS once attribution gets fuzzy. A MER of 3.0 means EUR 3 of revenue for every EUR 1 of ad spend; sustaining MER > 2.5 with healthy margin is the typical scale target.

P

Persistent ad An ad still running after a defined threshold — CommonWealth Ops's classification uses 30 days as the persistence threshold. Persistent ads are the strongest non-public-impression signal of profitability: an unprofitable ad gets paused; a persistent one is paying for itself. The classification is one of the trending_score inputs.

Pixel A tracking tag (Meta Pixel, TikTok Pixel, LinkedIn Insight Tag) installed on the brand's site to record visits, events, and conversions. The pixel feeds the platform's optimisation algorithm and is the source for custom audiences and lookalike seeds. Privacy regulation (GDPR, CPRA) has made pixel consent management increasingly load-bearing.

Problem-first hook A hook archetype that opens with the user's problem before introducing the brand or product. Example: "If your face wash leaves your skin tight, here's why." Problem-first hooks work for new categories and for buyers in active discomfort; they underperform for buyers in delight-purchase mode.

R

Reach The number of unique users who saw an ad at least once over the campaign window. Reach plus average frequency multiplies to total impressions. The public libraries do not disclose reach. Operators measure reach via their own ad account.

Result-based hook A hook archetype that opens with the desired result. Example: "Drop 5 pounds in 14 days without changing your diet." Result-based hooks work for outcome-driven categories (fitness, wealth, skincare) and underperform for categories where the result is hard to credibly promise.

Retargeting Showing ads to users who already engaged with the brand — site visitors, cart abandoners, past purchasers, video viewers. Retargeting is bottom-of-funnel by definition: it operates on a known-warmed audience. Strongest retargeting performance comes from segmented retargeting (cart abandoners get different creative than past purchasers), not from one-size-fits-all retargeting campaigns.

ROAS (Return on Ad Spend) Revenue divided by ad spend. ROAS is the per-campaign performance metric. ROAS of 3.0 means EUR 3 of revenue for every EUR 1 of ad spend. Because ROAS does not subtract product cost, fees, or fulfillment, comparing two ROAS numbers without comparing product margin and LTV is a common operator mistake.

Run duration The number of days an ad has been continuously running on the public surface. Run duration is one of the highest-signal observable proxies in ad intelligence — an ad still running at day 30+ is paying for itself. CommonWealth Ops's persistent-ad classification uses run duration as a primary input.

S

Scarcity trigger A copy device using limited supply (only 100 left, while stocks last) to compress decision time. Scarcity converts on undecided buyers and under-converts on buyers in research mode. Over-used scarcity erodes brand trust — operators rotating scarcity triggers across creative tend to avoid the trust decay.

Scroll stop The capacity of a creative to halt the user's scroll. Distinct from hook rate (which measures post-stop retention), scroll stop is the binary did-they-stop-at-all metric. Static ad scroll stop is driven by visual contrast and pattern interrupt; video ad scroll stop is driven by the first frame and motion design.

Social proof trigger A copy device using third-party validation (reviews, user counts, press logos, testimonials) to convince buyers. Social proof converts strongly when the proof is specific (named users, screenshot reviews) and weakly when generic (stars, vague claims). User-generated content is structurally social proof at the creative level.

Spend (estimated) Meta discloses ad spend ONLY for political ads in coarse bands (USD 0-100, 100-500, etc.). For commercial ads no spend data is exposed. Tools claiming to disclose competitor commercial spend are either inferring from third-party panel data with wide error bands or extrapolating from indirect signals. CommonWealth Ops does not claim competitor spend.

Static ad A non-video ad — image, carousel, collection. Static ads have lower production cost than video, lower engagement on TikTok, and competitive engagement on Meta when the image carries strong visual message hierarchy. Static-vs-video mix is platform-dependent.

T

TikTok Ad Library TikTok's public archive of ads, available at TikTok Creative Center. The library discloses advertiser identity, ad creative, run dates, target countries, and (for many ads) public view counts as views_band. The library is structurally less mature than Meta's but improves quarterly.

Trending score CommonWealth Ops's internal creative-ranking metric within one (niche, week, language) tuple. Inputs include run duration, advertiser concentration, hook archetype frequency, and trigger combination frequency. The score is relative within a niche-week, not absolute across niches. Documented in our pipeline post.

Top-of-funnel (TOFU) Awareness-stage content and ads targeting cold audiences. TOFU CPMs are lower than BOFU but conversion is delayed. Operators starving TOFU usually run out of audience at the bottom of the funnel within 6-9 months. TOFU content includes long-form posts, brand videos, and broad-audience ads.

Trigger combination The emotional levers stacked in a hook — scarcity, social proof, urgency, identity, result, problem. CommonWealth Ops classifies each captured ad by trigger combination. Niches dominated by single-trigger ads (only scarcity) underperform niches that mix two or three triggers per ad.

TTC (Time-to-Conversion) The number of days from first ad exposure to purchase. Median TTC for low-AOV impulse buys (under EUR 30) is often same-day or +1 day; high-AOV considered purchases (EUR 200+) can stretch to 14-30 days. TTC informs attribution window selection and retargeting cadence.

U

UGC (User-Generated Content) Content shot by creators or customers rather than by the brand's studio. UGC ads typically outperform polished studio ads on TikTok and increasingly on Meta because they read as authentic rather than promotional. UGC production cost varies from EUR 100-1000 per creative depending on creator scale.

Urgency trigger A copy device using time pressure (24 hours only, ending tonight) to compress decision time. Urgency works for impulse purchases and for buyers already in consideration; it under-performs for cold awareness. Same trust-erosion caveat as scarcity if over-used.

V

views_band The view-count range disclosed by TikTok for ads in its public library (e.g. 0-1K, 1K-5K, 5K-50K, 50K+). Not a precise impression count; a coarse bracket. CommonWealth Ops uses viewsband as one of the trendingscore inputs on TikTok ads, weighted lower than run duration because of the bracket coarseness.

View-through conversion A conversion attributed to an ad the user saw but did not click. Meta and TikTok report view-through conversions in the advertiser account with a configurable window (typically 1-day view). View-through conversions inflate apparent ROAS — operators reading ROAS as click-only see a more conservative number.

W

Whisper ASR Whisper Automatic Speech Recognition — the open-source speech-to-text model from OpenAI. CommonWealth Ops uses the faster-whisper community optimisation inside the weekly intelligence pipeline to transcribe spoken hooks of TikTok video ads whose hook_text the public library did not surface. The transcribed hooks become part of the searchable archive — meaning the spoken-only hook of a TikTok video is captured, not just the on-screen caption.

What this glossary deliberately leaves out

Three categories of term that some glossaries include but this one does not, and why.

Platform-specific UI labels. Meta and TikTok change UI labels (Ad Set vs Ad Group vs Campaign Group) on a quarterly basis. A glossary that tracks UI labels rots inside three months. Where the platforms changed terminology recently (Reach + Frequency to Awareness; Engagement Rate to multiple sub-metrics), we used the durable concept.

Vendor-specific jargon outside CommonWealth Ops. Other ad-intelligence tools have proprietary terms (AdSpy's smart filters, Minea's winning products). We defined our own (trendingscore, viewsband as we surface it, Whisper ASR pipeline). For competitor terms, the competitor's docs are the authoritative source.

Conversion-rate-optimisation terms. Landing-page CRO is its own field with its own glossary — heatmap, scroll depth, time-on-page, micro-conversion. They belong in a CRO glossary, not an ad-intelligence glossary.

How to use this glossary

Bookmark it. Send the URL to junior team members. Use it as the canonical reference when team members use a term differently — a shared vocabulary moves decisions faster than re-defining terms in every meeting. We will update this list quarterly as the field evolves; meaningful changes will be noted in the post's last_updated field.

For context on how these terms appear inside CommonWealth Ops's weekly intelligence, read our pipeline post and our scope-honesty post.

See pricing →