Inside the campaign · Measurement & buying
The useful question is not just “How many ads ran?” It is what the available evidence helps us decide about your message, placements, follow-up and next investment.
TV devices
The annual record contains 461,275 impressions classified as TV-device delivery and 445,307 recorded completions. Completion divided by starts was about 97.0%.
Playback completion is not proof of attention, unique household reach or sales. Device labels alone do not verify a specific publisher, show or FAST placement.
Interactive devices
Outside the two known retargeting lines, phones, tablets and computers recorded 1,918 clicks from 1,704,193 impressions. We review those responses alongside available website activity and confirmed business outcomes.
A click is a diagnostic signal. It is not automatically an engaged visit, a qualified inquiry or a customer.
Branding and retargeting have different jobs. Their audiences, repetition and creative should be planned deliberately. Available identifiers, privacy choices and platform support affect who can be reached and how repetition can be managed.
Takeaway: matching coverage needs investigation before attributing the decline to demand. Partial August/October periods; these event counts are not a click-to-visit conversion rate. Source: Pontiac journey export, conversion dates Aug 2–Oct 30, 2026; reviewed Nov 1, 2026.
The recent journey export associated 94 recorded visit-event IDs with prior advertising exposures. Eighty-five paths included a click flag; nine did not. These are matched events—not 94 identified people or proven incremental visits.
Matched visit IDs fell from 82 in August to 11 in September and 1 in October, even while advertising continued. The October delivery record includes 252 clicks.
What Rock does with that: prioritize a tracking and matching review before calling it a change in customer demand or moving money based on attributed results.
The inquiry-path export contains no linked ad-exposure rows. That does not establish zero inquiries. It means this report cannot identify which ad produced a qualified lead.
What Rock does with that: reconcile successful-form tracking with lead and sales records before making a revenue claim.
Conversion-date period: August 2–October 30, 2026; 30-day attribution lookback. August and October are partial months. Matched-event counts and delivery clicks are different populations, so their ratio is not a conversion rate. IP-based associations do not identify a specific person or establish causation.
Inspect exact domains, app identifiers, supply authorization and useful on-site activity. A familiar publisher name or a high CTR does not settle quality.
Review campaign and line controls together. Distinguish a device ID from a shared IP and investigate repeated exposure before promising household-level limits.
Compare similar-length ads in comparable delivery conditions. Use completion and clicks for their appropriate roles rather than naming a winner from one headline metric.
Choose one question, preserve the control and agree on a useful outcome. Make a change we can evaluate instead of changing every variable at once.
For your business
This is how we approach our clients’ advertising, too: connect the business goal, commercial, media plan and available response data. We agree on the reporting, tracking and production scope around your campaign goals and investment.
Rock brings video production and campaign management together for businesses in New Hampshire, Massachusetts, Maine and Vermont. Humans lead the strategy, production and buying decisions; AI assists analysis and the development of test ideas.