Comparing Ad Platforms? Five Questions to Ask Your Measurement Tool First.

Nov 18, 2026
Armanet
Armanet

Comparing Ad Platforms? Five Questions to Ask Your Measurement Tool First.

Armanet brings modern targeting, attribution, and reporting to the firearms and outdoor space, while Triple Whale, Northbeam, Rockerbox, and others expand what marketers can measure. Better data is welcome, but more dashboards don't necessarily mean clearer answers.

Armanet, GA4, and Shopify can tell different stories about the same sale. All three may be right, because each tracks customers, counts conversions, and assigns credit differently. These five questions help you compare like with like rather than mistaking measurement differences for advertising performance.

1. Does Your Measurement Tool See Impressions or Only Clicks?

Most buyers don't see one banner ad for a $2,000 optic and immediately buy. They read reviews, watch videos, and compare products, seeing your ad five or six times before searching Google. Click-only reporting makes Google look like the whole story, overlooking the advertising that came before it.

Some tools track only clicks; others connect impressions with later purchases. Armanet brings that broader visibility to the firearms industry, linking eligible impressions with downstream conversions, including purchases from shoppers who never clicked.

Why View-Through Conversions Matter

Armanet reports view-through conversions (purchases within the attribution window by people who saw but never clicked an ad) alongside click-through conversions. Neither report proves causation; together they show exposure and direct response.

NERD TIME

Clicks Were Just the Tip of the Iceberg

What the Yahoo study found

A randomized experiment with 1,577,256 retail customers found a 0.28% click-through rate; only 7.2% of ad viewers ever clicked. Yet 78% of total sales lift came from people who saw ads but never clicked (Lewis & Reiley, 2014).

On the Armanet network

Our multitouch analysis of 1,285 purchases across seven anonymized firearms and outdoor brands found 80% of buyers saw multiple Armanet ads, averaging five over three days. One $800 optic buyer never clicked the first ad, returned 16 hours later, saw eight more, and purchased two days after the first impression.

The Second Ad Matters. So Does the Fifth.

A separate analysis of 15 Armanet advertisers found 52% of conversions followed repeat exposure, meaning the second impression or later, and all 15 had at least 30%.

Ad exposureCumulative households converting
After the 1st ad1.0x baseline
By the 5th ad4.5x the first-ad total
By the 10th ad7.1x the first-ad total

These are cumulative household conversion counts, not evidence an individual household became 7.1x more likely to convert after ten ads. Gains slowed around ten impressions but continued. Successive ads contributed at declining rates relative to the first:

  • 2nd ad: Approximately 0.9x the first ad's contribution.
  • 3rd through 5th ads: Approximately 0.6x to 0.8x.
  • 6th through 12th ads: Approximately 0.35x to 0.5x.
  • 13th ad and beyond: Approximately 0.3x.

New households showed the same curve, so the pattern wasn't simply retargeting existing customers. Methodology matters: this observational, non-randomized analysis adjusted for browsing intensity but doesn't prove each impression caused a purchase. Unconnectable cross-network purchases understate conversion rates, making the per-ad decline conservative.

What the measurement tools say

Triple Whale captures views from Meta, TikTok, and Pinterest; Northbeam adds Snap and more; Polar captures Facebook views. Other channels may rely on clicks and UTMs, and Northbeam notes click-only measurement favors campaigns nearer the final purchase. A click-only tool may record the sale but cannot credit earlier, non-clicked impressions.

Sources: Lewis & Reiley (2014), Online Ads and Offline Sales; Armanet Multitouch Case Study; Triple Whale; Northbeam; Polar.

2. Which Attribution Model and Window Is It Using?

Even tools that see the same interactions can assign credit differently. A shopper who sees a display ad Monday, clicks a Google ad Wednesday, and buys Friday could be credited to the last click, the last advertising channel, or several interactions.

The attribution window determines how far back each tool looks, and a one-day window and a month-long window won't describe the same journey, especially for a premium rifle versus a box of ammo. The goal isn't one model with every answer; it's understanding the model behind each report. Share your model and window with advertising partners so it's clear why a sale appears in one report but not another.

NERD TIME

The Ads Kept Working After the Campaign Ended

What the Yahoo study found

A two-week campaign produced R$0.102 in sales lift per person, approximately 5%. Cumulative lift reached R$0.166 after three weeks and R$0.298 by eight weeks, with a significant effect the week after the campaign ended (Lewis & Reiley, 2014). Short windows can miss purchases influenced by earlier advertising.

On the Armanet network

In our multitouch study, 73% of purchases occurred at least one day after the first ad; a one-day window would capture only 27%. Orders over $500 had a median of seven ad touches, versus four for smaller orders.

What the measurement tools say

PlatformAttribution settings or window
GA490-day default lookback for most key events
ShopifyLast non-direct click by default, 30-day path view
Triple Whale28-day click, 7-day view
Rockerbox150-day click, 90-day display view, 1-day modeled view
Armanet case study30-day click, 14-day view

These settings describe different reporting methods and may be configurable, so the same purchase can produce different answers under different windows.

Sources: Lewis & Reiley (2014), Online Ads and Offline Sales; Armanet Multitouch Case Study; GA4; Shopify; Triple Whale; Rockerbox.

3. Is Every Platform Being Tracked the Same Way?

Tracking differences can start before a customer buys. One platform's tag may load immediately while another waits for cookie consent. UTMs, the labels added to campaign links, also depend on consistent setup. Review tags, conversion events, consent settings, and campaign labels across platforms; knowing what gets recorded makes discrepancies easier to explain.

NERD TIME

Sloppy Tracking Makes Ads Look Weaker

What the Yahoo study found

Researchers randomly assigned 81% of 1,577,256 matched customers to advertising eligibility and held out 19% (299,426 people) as controls. Only 63.7% of eligible customers actually saw an ad (Lewis & Reiley, 2014). Defining eligibility separately from exposure kept researchers from treating every eligible shopper as an ad viewer.

On the Armanet network

Consistent event definitions and tracking practices help advertisers compare impressions, clicks, and conversions meaningfully.

What the measurement tools say

Google Campaign Manager 360 says discrepancies of up to 20% between ad servers can be normal; date ranges, edited or duplicate tags, invalid-traffic filtering, and consent settings affect counts. Give vendors consistent tracking conditions before reading differences as performance gaps.

Sources: Lewis & Reiley (2014), Online Ads and Offline Sales; Campaign Manager 360 Help; Google Analytics Help; Shopify Help Center.

4. How Does It Know It's the Same Person?

A shopper may discover a product on their phone, research it on a laptop, and buy from a desktop: one journey, but some tools may see three shoppers. Cookies, logins, device identifiers, and proprietary identity technology connect those interactions differently.

Armanet uses ArmanetID, anonymized intelligence infrastructure built for the firearms and outdoor industry. It connects eligible advertising interactions and conversions across our network without third-party cookies or direct personal identifiers like names or email addresses. The useful question isn't how many visitors a tool reports, but which interactions it can reliably connect.

NERD TIME

The Whole Experiment Ran on an Identity Match

What the Yahoo study found

Researchers matched 1,577,256 retailer customers to Yahoo accounts using names and either email or postal addresses, linking ad exposure to online and in-store purchases. A third party later anonymized the matched dataset (Lewis & Reiley, 2014).

On the Armanet network

ArmanetID derives anonymous identifiers from technical signals to support targeting, retargeting, and attribution. It doesn't collect names, email addresses, physical addresses, or phone numbers.

What the measurement tools say

Shopify uses device cookies, so phone and laptop visits can count separately. GA4 models certain activity when consent prevents persistent identification; Polar connects journeys through logins and digital fingerprints. Different identity methods produce different visitor counts for the same shoppers.

Sources: Lewis & Reiley (2014), Online Ads and Offline Sales; ArmanetID; Shopify Help Center; GA4 Help; Polar.

5. Does Your Measurement Tool See Where Customers Actually Buy?

For many firearms and outdoor brands, discovery and purchase happen in different places: a shopper might see your ad, research on your website, and buy from a favorite retailer, a sale Shopify or GA4 may never see.

Armanet's Halo reporting connects eligible transaction data from participating retailers' websites with advertising across our network, giving brands visibility beyond their own ecommerce stores. Our Halo reporting article explores why that matters for retailer-heavy brands.

NERD TIME

93% of the Lift Happened in Stores

What the Yahoo study found

Over three weeks, estimated lift was R$0.166 per person: R$0.155 in physical stores and R$0.011 online. Approximately 93% of the advertising effect happened offline, versus roughly 85% of normal sales (Lewis & Reiley, 2014). Online-only measurement would have missed nearly all the impact.

On the Armanet network

The Halo study covered five brands in July and August 2026 with $274,000 in ad spend. Brand websites showed 5.7x ROAS; tracked online retail sales ($1.83 million versus $1.55 million on brand websites) raised it to 12.4x, or 10.2x counting only incremental retail sales. Participating retailers represented approximately 15% of ecommerce, and no physical in-store sales were included, unlike the Yahoo study.

What the measurement tools say

Rockerbox connects retail transactions to advertising when a matching email or user ID is available at purchase. Shopify and GA4 measure purchases through configured stores and integrations. Retail visibility requires the right data connections, not just a website report.

Sources: Lewis & Reiley (2014), Online Ads and Offline Sales; Armanet Halo Case Study; Rockerbox Help Docs.

Bringing It Together: Attribution vs. Incrementality

Attribution describes the ads someone interacted with before buying. Incrementality asks whether they would have bought anyway. Holdout testing compares groups eligible and ineligible for advertising to estimate the additional sales a campaign generated.

Armanet offers this through Armaview for qualifying campaigns, with more detail in our incrementality article. Attributed sales are purchases connected to advertising under a tool's rules; incremental sales are the additional purchases advertising generated. Keeping that distinction clear prevents treating every observed conversion as a sale that would never have happened.

NERD TIME

The Shortcut That Got the Answer Backwards

What the Yahoo study found

Simply comparing ad viewers with non-viewers would have suggested ads reduced sales by R$0.23 per person. Heavy Yahoo browsers already spent less, creating bias opposite in direction and more than three times the true effect; randomized controls revealed the actual positive impact (Lewis & Reiley, 2014). Even with nearly 1.6 million customers, widely varying spending gave only a 63% chance of detecting a campaign that doubled its money.

On the Armanet network

Halo tracked $1.83 million in retail sales among ad viewers; holdout analysis estimated $1.24 million was incremental, approximately two-thirds of attributed sales. Armaview requires enough campaign scale. Frequency analysis shows how conversions accumulate with exposure but can't establish which purchases ads caused. That is another reason the two methods complement each other.

What the measurement tools say

Northbeam distinguishes causation from association and cautions that small audiences or short tests complicate interpretation. Northbeam, Rockerbox, and Triple Whale offer incrementality or lift testing alongside attribution.

Sources: Lewis & Reiley (2014), Online Ads and Offline Sales; Armanet Halo; Armanet Incrementality; Northbeam.

Your Measurement Checklist

Before comparing platform results:

  • Establish your reporting reference. Share your measurement tools with advertising partners.
  • Understand impression visibility. Check whether tools capture views or mainly clicks.
  • Review attribution settings. Compare models and windows.
  • Keep tracking consistent. Coordinate tags, conversion events, and consent settings.
  • Check UTMs. Verify campaign, source, and medium labels.
  • Include retail visibility. Look beyond your own website.
  • Understand incrementality. Use holdout testing to complement attribution.

More data is useful when you know what it represents. Website analytics, impression reporting, and retail measurement each reveal different parts of the journey. Armanet brings them together for the firearms and outdoor industry, from the first impression to the final purchase.

Connect With the Armanet Team

Note on the Yahoo study: The retailer disguised its actual financial figures, so the researchers report amounts in R$ (retail dollars). These figures represent relative magnitudes rather than literal US dollar amounts.

Related Articles

All posts
Top