Site index
Everything on this site, in one directory. The same list is linked from the footer of every page.
Everything on the site
Pages
- Growth partner for paid acquisition, built for the AI era of searchThe operator behind your paid acquisition: paid search, answer engine advertising, measurement and agent systems, run as one operation.
- How acquisition gets run, and what gets built around itThe relevance model, what the work focuses on, what gets built, and the four capability pages.
- Working notes on where search and marketing are heading with AIResearch notes that argue something, grouped by topic: paid search, answer engines, measurement, website relevance and local search.
- A growth practice built on how the platforms actually workWho Digital Traction is, and what an engagement looks like.
- What happens if you replyThe three stages of an engagement, and the one form on the site.
- Future of SearchHow search, answers, advertising, websites and agents are changing: what is durable, what is moving, and what is still uncertain.
Capabilities
- Answer engine advertisingSponsored placement in ChatGPT and on Reddit first, what the search account already earns inside Google and Copilot, and the emerging networks watched until a test can be read.
- Paid searchThe query, the ad, the page and the signal should feel like one conversation.
- MeasurementAutomation optimizes toward whatever you count. A form fill is the easiest thing to count.
- AgentsNothing reaches a live account or a live site without someone opening the door for that specific change.
Research notes
- What a growth partnership with Digital Traction looks likeAn example of the work, based on active client work and anonymized.
- Approved is not servingGoogle is extending impression caps based on account trust signals to every Ads product, and because nothing gets disapproved it looks like weak performance rather than a policy state
- Campaign-level lift is not account-level liftThe platform's own performance figure is measured per campaign, which is the level at which reallocated conversions look identical to new ones
- Google stopped letting your campaigns beat their targetBudget-limited campaigns used to quietly outperform their bid target, and that has stopped, which raises cost per lead in every account where the target was never really a decision
- The August spam update changed enforcement, not the rulesNo new policies and no blog post, which means the rules held and the enforcement of them improved, and that is the harder version to be caught by
- Answer-engine visibility does not have a scoreboard yetEvery available source of AI visibility data is partial in a different direction, and the responsible move is to say so rather than to average them into a score
- Business Agent for Leads changes what arrives in your CRMGoogle is replacing the lead form inside the ad with a Gemini chat grounded in your own site, which moves qualification in front of the conversion and breaks the metric most lead-gen accounts are managed on
- Four Google controls that people keep mistaking for each otherGoogle-Extended, the Search generative AI control, nosnippet and noindex sound like versions of the same switch and each one does something different
- Google's AI-search guidance is mostly things to stop doingGoogle published official guidance on generative AI features and the substance of it is a mythbusting list, which is more useful than the reassurance it came wrapped in
- Local Services Ads are moving into Google AdsThe standalone LSA dashboard is being retired and campaigns migrate to a pay-per-lead Performance Max type, keeping the lead model but removing manual bidding, vertical-level targets and your historical reports
- When a location page becomes doorway abuseGoogle's spam policy names city-targeted pages that funnel users to one destination, which describes the standard multi-location page set almost exactly
- Almost nothing requests your llms.txt fileGoogle says Search ignores the file, no major provider has committed to reading it, and a study of 38,000 domains carrying one found almost none were ever requested
- Query fan-out does not mean you need more pagesThe model generates related queries around yours before answering, and the common response of building a page per variation is the one thing Google explicitly warns against
- Why healthcare ads serve less than their settings implyGoogle labels many healthcare ads Eligible (limited), which is not a disapproval and does not appear as an error, so most practices experience it as unexplained low volume
- What Search Console's AI report tells you, and what it leaves outThe generative AI performance report gives you impressions from AI Overviews and AI Mode and nothing else, which makes it a visibility signal rather than a performance metric
- When you cannot retarget, what you count becomes the whole strategySensitive-category rules remove the audience levers most accounts lean on, which leaves the conversion definition as the main thing an operator still controls
- Where AI traffic actually lands in GA4GA4's AI Assistant channel covers ChatGPT and its peers but explicitly excludes AI Overviews and AI Mode, which sit in Organic Search, so there is no single AI number to report
- Where an AI answer gets its picture of your practiceGoogle's own guidance points local businesses at Business Profile and Merchant Center rather than the website alone, which means the answer is assembled from sources most practices never check
- Your campaigns are being upgraded to AI Max in SeptemberAutomatically created assets and campaign-level broad match flip to AI Max from September 2026, which widens matching and hands text and landing-page choices to Google unless you set the controls first
- What an agent can determine from a business websiteA running test of what a browser agent can reliably determine from a public business website
- Answer-first page openingsA planned test of whether opening a page with the direct answer changes retrieval and citation
- Does anything fetch llms.txt?A running fetch log testing whether any user agent actually requests this site's llms.txt
- Read versus Inspect retrievalA running comparison of what a reader sees against what a machine extracts from the same page
- Internal-link relationship labelsA running test of whether labelling the relationship an internal link expresses changes how it is interpreted
- What answer-engine advertising actually is todayIt now spans two different models, existing ad systems extending into AI search and purpose-built conversational ad platforms, and the honest thing is to keep them apart
- AI answers are changing what a click is worthWhen an AI answer resolves the question on the results page, the clicks that still come through are fewer but more decided, which changes how you should value and bid on them
- Feeding real outcomes back to the platformForm fills teach automated bidding the wrong lesson; importing qualified leads and revenue from your CRM teaches it what a real opportunity looks like
- Reddit is where high-consideration buyers compareFor considered purchases, people go to community threads to sanity-check their options, which makes Reddit a discovery surface worth reading, and a place to buy ads
Guides
- Connecticut Google Ads consulting for lead-generation businessesSenior, operator-led Google Ads consulting for Connecticut lead-generation businesses, focused on qualified leads and cleaner measurement, available nationally
- Find out where your Google Ads account is losing efficiencyAn independent, lead-quality-led review of where a Google Ads account loses efficiency
- Give Google Ads better information about which leads matterA two-layer conversion-tracking and CRM-signal setup that teaches Google Ads to bid on qualified leads, not raw form fills
- Running Google Ads for an aesthetics practiceWhat actually governs paid search for an injectables or cosmetic surgery practice, and the order to fix it in
- Running paid search for a home services business, mid-migrationWhat the Local Services Ads migration into Google Ads changes for a trades business, and the order to handle it in
- Google Ads for lead generation should optimize beyond the form fillHow Google Ads lead generation should optimize toward qualified customers, not just form fills
- What Google's health targeting rules actually forbidWhich Google Ads targeting features are unavailable to practices advertising injections or cosmetic surgery, and what remains
- Why Google Ads can generate conversions but still produce low-quality leadsWhy Google Ads can report conversions yet produce low-quality leads, and how to diagnose the media, website, and sales causes
- Hands-on paid search consulting built around lead qualityHands-on, operator-led paid search consulting built around lead quality, not clicks
- The Digital Traction paid search diagnosticThe fixed-order method Digital Traction uses to read a Google Ads account, measurement first
- The qualified-lead feedback loopHow Digital Traction returns CRM outcomes and revenue to the ad platform so bidding buys the clicks that become customers
Reference
- GlossaryThe vocabulary this site uses precisely.
- Sources and methodThe source register and method behind the site: claim types, confidence labels, checked dates, primary-source preference, and unverifiable handling.
- Site indexEverything on this site, in one directory.
- Agent indexStructured facts about this site and organization, for people, crawlers and software agents alike.
- What you can actually buy on AI search surfacesA maintained record of paid placement across AI search and answer engines: what is buyable today, what is in testing, and what cannot be verified.
Glossary terms
- AEO (answer engine optimization)Work that makes content more likely to be retrieved, interpreted correctly and cited by answer engines. The work is structural: answers, evidence, entities and dates, which is a wider job than chasing rankings.
- Agent-ready websiteA website whose structure, labels, states and actions are legible and completable by software acting on a person's behalf, achieved through the same qualities that make it accessible to people.
- Answer engineA system that composes a direct response to a question from retrieved and interpreted sources, rather than returning a ranked list of links.
- Answer-engine visibilityThe observable presence of a business in machine-composed answers: citations, mentions and links inside AI Overviews, assistant responses and similar surfaces.
- Brand evidenceThe consistent, verifiable body of material that lets a person or machine understand and trust a business: definitions, people, credentials, claims with proof, reviews, references, original research.
- Claims ledgerA maintained register of the claims a business makes publicly, each with its type, supporting evidence, sources, confidence and checked date.
- EntityA distinct, identifiable thing (an organization, person, service, place or concept) that search and answer systems try to recognize and connect across sources.
- Intent taxonomyA working map from real customer questions to the pages, ads, evidence and actions that should answer them, organized by what the asker is trying to do.
- llms.txtA proposed convention (Answer.AI, September 2024) for a markdown file at /llms.txt listing a site's key content for language models; no major AI platform has confirmed consuming it.
- Offline conversion importReturning conversions that complete outside the website (qualified opportunities, consultations, revenue) to an ad platform, matched by click identifier or hashed customer data.
- Query fan-outThe expansion of one user question into many related sub-queries by a search or answer system, which then retrieves against that whole expanded set, reaching well past the literal words typed.
- Signal qualityHow accurately and completely the events returned to an optimizing system represent the outcomes the business actually values.
The same read, on your account
The diagnosis is the first step and it is not charged for. What is offered is an honest read and sensible, evidence-backed decisions, not a promise about lead volume.