Curtis Howland shares paid-advertising and growth advice for premium ecommerce brands, including Meta and Google ads, AI search visibility, and campaign measurement.
Curtis Howland is most useful for paid-ad and AI-search operating advice for premium ecommerce growth teams.
Best for
Premium ecommerce brands or DTC agencies hiring a paid-ads and AI-search operator.
Not for
Mass-market, low-AOV, impulse-led brands paying for consumer reach via sponsored videos.
Key takeaways
AI-search audit uses 20 to 30 real buying questions in the shopper's wording.
HexClad gets roughly half its revenue in November and December.
Click tracking and surveys disagree about AI's role in sales.
Check first: Confirm agency versus sponsorship and ask for audience demographics and post results.
Main ideas
Click tracking misses most AI-influenced revenue · Answer-engine optimisation is an operator checklist, not a mystery · Rebuild creative for the placement; don't reuse it unchanged
Curtis Howland shares practical growth advice for premium ecommerce brands, especially on paid advertising, AI search visibility, and measurement.
Their posts turn marketing decisions into specific operating steps: for example, they explain how to test shoppers’ AI-search questions and how HexClad plans advertising around a long buying season.
They also cover ad creative for mobile-game placements and a DTC valuation calculator.
Howland’s work is strongest when it links a marketing choice to the mechanics behind it.
In their HexClad breakdown, they connect a long consideration period and concentrated holiday sales to lower October efficiency targets and planning well ahead of the sale.
In their AI-search guide, they give a concrete audit process based on real buying questions, credit Lorea Lastiri for the playbook, and say they are not an AEO expert.
That qualification matters: the post offers a practical guide, not specialist authority.
Across the examples, readers get useful questions, measurements, and test plans—not just channel predictions.
The posts report outside study findings and business figures, which were not independently checked here.
Curtis Howland's posts are written for premium ecommerce brands whose customers take months, not seconds, to decide.
Read more
Two priorities run through them.
One is measurement honesty: Howland argues that click tracking badly understates how much revenue AI assistants now influence, and that the fix starts with a single question added to a post-purchase survey.
The other is channel mechanics: whether the channel is AI answers, in-game mobile ads, or a November sale window, Howland's argument is that the specifics of the format decide what a brand should build, budget and measure.
The work blends agency positioning with a founder-facing exit frame, and it consistently teaches through named third-party data and operator quotes rather than opinion.
Click tracking misses most AI-influenced revenue
Howland's measurement argument is that assistants send shoppers to a site without a click, so UTM and pixel attribution systematically undercount AI's contribution and surveys catch what clicks miss.
Read more
Across Fairing's tracking of 158 DTC brands spending $10M to $2B a year over 104 weeks, post-purchase surveys caught 7.5 times more AI-influenced orders than click-based attribution and 10.6 times the revenue, with click methods capturing under 10% of the AI-driven revenue surveys found. Howland reports AI's share of gross merchandise value growing 12 times year over year, and argues the category you sell into changes how much this matters.
High-consideration, high-value categories such as mattresses and health devices show the largest AI-driven shares, while in commoditised categories AI agents default to one or two known brands and smaller brands drop out of consideration entirely.
In housewares the spread between the most and least AI-attributed brand widened from about two points to more than twenty in a year.
The practical fix Howland recommends is one added survey question asking whether an assistant helped the shopper find the brand, with a threshold of 100 or more responses at a 30% or higher response rate before acting on the split.
Howland notes a reversal: a year earlier, AI-referred visitors converted at roughly half the rate of other traffic; now, the reported figure is 54% higher.
Answer-engine optimisation is an operator checklist, not a mystery
Howland lays out ten concrete tasks for brands that have not started working to get cited by AI assistants.
Read more
Start by harvesting 20 to 30 real buying questions from support tickets, reviews, site search and post-purchase surveys, in shoppers' own words.
Enter each question word for word into ChatGPT and Google AI Mode repeatedly, logging the answers, whether the brand appeared and which sources the answer pulled.
Then address the underlying conditions: keep price, availability, variants, shipping and returns in visible HTML so the Merchant Center feed matches the page; publish data only the brand holds, such as return reasons and sizing failures; and write comparison pages that name who should not buy.
The post distinguishes OpenAI's three crawlers by function: GPTBot for training, OAI-SearchBot for search inclusion and ChatGPT-User for user-triggered fetches.
It notes that robots.txt changes take roughly 24 hours to register on OpenAI's side.
It also advises deep-linking citations to the relevant product page rather than the homepage: a great ad pointed at a generic page underperforms a mediocre one pointed at the right page.
Rebuild creative for the placement; don't reuse it unchanged
Howland argues that the ad placement's interface dictates the edit.
Read more
In-game ads on AppLovin behave differently from a Meta feed: rewarded video locks viewers in for 15 to 30 seconds in exchange for in-game currency, while interstitial ads show a three-second unskippable window before the Skip button appears. A concept proven on Meta must be recut around the placement's clock, not dropped in unchanged, and tested as a separate format.
For rewarded video the suggested structure is story or demo through second 25, tension through second 28, and the strongest call to action in the final seconds before the viewer can close it.
For interstitials the hook should land at second three, the moment Skip becomes tappable, which Howland calls the opposite of Meta, where the ad interrupts a scroll and must hook in the first three seconds.
The cheatsheet is to take the three best-performing Meta concepts, recut the timing, run rewarded and interstitial as separate tests, and judge the channel on blended marketing efficiency ratio because in-app attribution is messier than Meta's.
Howland says that a few years earlier the idea of putting a $60 physical product in the middle of someone's mobile game would have been laughable, and now frames these networks as performing well for premium ecommerce, especially brands selling to women aged 40 to 50.
Premium brands should fund the months when nobody is buying
Howland argues that for products with a long consideration window and a concentrated buying season, flat months are when demand is built; below-target returns then are a deliberate choice, not a failure.
Read more
HexClad Cookware makes roughly half its revenue in November and December, yet runs advertising from January to October at low return. October runs at a marketing efficiency ratio about 33% below the year-end target, and the 13 days before the sale run about 50% below it.
Under 20% of purchases happen inside a seven-day window, and 80 to 85% of buyers say they knew the brand for months first, against a $400 average order value.
Black Friday weekend has the lowest return lift of the whole period; the lift peaks at over 300% on 12–13 November, before the sale opens.
HexClad's president says the company left about $1M of Black Friday budget unspent because the bottom-of-funnel work had already been done.
Offers, timing and targets are locked about 15 weeks out.
Growth work should be readable through an exit lens
Howland frames part of the advice for founders rather than media buyers, and offers a DTC valuation calculator to a founder audience, introduced by the claim of having helped five DTC brands sell for $500M or more.
Read more
The described tool covers a valuation engine that turns a profit and loss statement into adjusted EBITDA and low, base and high multiples; a multiple builder with eight levers, each priced in turns; the bridge from enterprise value down to the cash that actually wires, accounting for debt, working capital, inventory, fees, rollover and tax; a cost-of-capital normaliser that converts any flat fee or factor rate to a true annual rate and prices it against nine financing options; and a benchmarks tab with sourced multiple bands.
A worked example runs a $12M brand through the tool and reports $9.7M enterprise value and $5.6M cash at close. The post invites readers to request the calculator through a connection or comment rather than linking it directly.
The promotion introduces exit terms and financing literacy before the wider name search and the creator's campaign-fit context. It shifts the same audience of DTC ad buyers into the role of owners weighing a sale, bringing valuation scope alongside ad-channel advice.
Teaching through named data, operator quotes and a resource handoff
A recurring priority in how Howland writes: a post opens with a number or a named brand, moves through numbered or arrowed steps, and closes with either a question inviting comparison or an offer of a downloadable asset.
Read more
Depth comes from specifics rather than opinion — channel mechanics, dollar thresholds, percentages — and from quoting the operators being written about rather than paraphrasing them.
The AI attribution post cites Fairing, Adobe, Podscribe and a Columbia/Yale study by name, and asks readers whether AI mentions are climbing in their survey responses.
The HexClad breakdown quotes the company's head of growth and president directly.
Two of the five posts offer resources—a creative strategy guide and the valuation calculator—in exchange for a comment or connection request.
Answers a question premium brands face in their first autumn: if almost nobody is shopping for your product today, what should your ad account be doing?
Read more
This is the set's most concrete example, with numbers attached to each decision.
A breakdown of HexClad, the hybrid cookware brand, and why it deliberately runs October below target.
The post says roughly 50% of revenue lands in November and December, with 12% on Black Friday alone, against a $400 average order value.
Under 20% of purchases happen inside a seven-day window, and 80–85% of buyers say they knew the brand for months first.
October runs at a marketing efficiency ratio about 33% below the year-end target; the 13 days before the sale run about 50% below it.
ROAS lift peaks over 300% on 12–13 November, before the sale opens, while Black Friday weekend has the lowest lift of the period.
HexClad's president says the brand leaves about $1M of Black Friday budget unspent because the bottom-of-funnel work is already done.
Offers, timing and targets are locked about 15 weeks out.
For brands planning a campaign
For a brand with a long consideration window, judge pre-season months by demand built, not last-click return, and lock the offer calendar early enough for creative and budget to follow. The 15-week planning horizon and deliberate October under-spend are the choices to copy or explicitly reject.
Explains what “AEO” means in day-to-day ecommerce work.
Read more
For readers who know the acronym but want an operating checklist, this is the clearest starting point in the set.
A numbered playbook for getting an ecommerce brand cited by AI assistants.
It distinguishes OpenAI's three crawlers — GPTBot for training, OAI-SearchBot for search inclusion and ChatGPT-User for user-triggered fetches — and notes that robots.txt changes take about 24 hours to register.
It recommends pulling 20–30 buying questions from support tickets, reviews, site search and post-purchase surveys in shoppers' own wording, then entering each into ChatGPT and Google AI Mode and logging whether the brand was mentioned, cited and which sources were pulled.
Other steps include putting price, availability, variants, shipping and returns in visible HTML while keeping the Merchant Center feed aligned with the page; publishing data only the brand has, such as return reasons and sizing failures; writing honest comparison pages with a dated methodology and named criteria; keeping product names consistent; correcting wrong facts on review sites and forums; deep-linking citations to the relevant product page; and tracking AI referrals, citation rates and AI visibility in Search Console.
The post credits the playbook to Lorea Lastiri and says the author is not an AEO expert.
For brands planning a campaign
Treat AI search as a scoped workstream with its own audit and measurement, not a footnote to a paid media retainer. The question set, crawl-eligibility check and citation log are small enough for a first deliverable before any larger commitment.
Answers the measurement question brands ask before investing in AI visibility: is it moving revenue, and how would we know?
Read more
It also explains why two credible methods can report very different numbers.
A summary of a Fairing study covering 104 weeks from November 2024 to May 2026 across 158 DTC brands spending $10M to $2B a year.
The post says post-purchase surveys caught 7.5 times more AI-influenced orders than click-based attribution, the revenue gap was 10.6 times, and click methods captured under 10% of the AI-driven revenue the surveys found.
It reports that AI's share of gross merchandise value grew 12 times year over year, Adobe tracked AI referrals to US retail sites up 14 times since October 2024, and AI-referred visitors now convert 54% better than non-AI traffic — a reversal from a year earlier.
High-consideration, high-average-order categories such as mattresses and health devices show the largest AI-driven shares.
In commoditised categories, AI agents default to one or two known brands.
In housewares, the gap between the most and least AI-attributed brand widened from about two points to more than twenty in a year.
The post recommends adding one post-purchase survey question about AI discovery and waiting for 100+ responses at a 30%+ response rate before acting on the split.
For brands planning a campaign
Add the AI discovery question to existing post-purchase surveys before funding anything larger, and set the sample threshold in advance so the result is not read before it is stable. In commoditised categories, presence among the small set of brands an assistant names is the deciding factor, not incremental reach.
Shows what changes when a brand tries a new ad channel.
Read more
It is the set's most format-specific example, demonstrating how placement mechanics, not just audience, drive the edit.
A case for testing AppLovin, which places full-screen ads inside mobile games and apps, for premium ecommerce brands.
The post describes two formats.
Rewarded video holds viewers for 15–30 seconds in exchange for in-game currency; the suggested structure is story or demo through second 25, tension through second 28, and the strongest call to action in the final seconds.
Interstitial ads show a three-second unskippable window before a Skip button appears.
The post argues the hook should land at second three, when Skip becomes tappable, rather than at second zero — the opposite of Meta, where the first seconds interrupt a scroll.
It claims the channel is performing for premium ecommerce, especially brands selling to women aged 40 to 50, and advises judging it on blended marketing efficiency ratio because in-app attribution is messier than Meta's.
The post offers a creative strategy guide in exchange for a comment.
For brands planning a campaign
Budget for a separate creative recut and separate tests per placement rather than reusing Meta cuts unchanged.
Agree in advance to judge the channel on blended efficiency instead of in-app attribution.
Treat the named audience segment as a testable hypothesis, not a given.
It connects the growth advice to an exit framing and gives founders a concrete tool description to evaluate.
A post offering a DTC valuation calculator.
Its components include a valuation engine that turns a profit and loss statement into adjusted EBITDA and low, base and high multiples; a multiple builder listing eight levers, each priced in turns; a bridge from enterprise value to cash at close covering debt, working capital, inventory, fees, rollover and tax; a cost-of-capital normaliser that converts flat fees and factor rates into a true annual rate and prices them against nine financing options; and a benchmarks tab with sourced multiple bands.
A worked example runs a $12M brand through the tool and reports $9.7M enterprise value and $5.6M cash at close.
The post invites readers to request the calculator.
For brands planning a campaign
Useful when a brand wants growth work framed around eventual sale value rather than short-term return, or needs to size how financing terms affect what actually wires at close. The post describes the tool but does not link it; ask for the calculator before relying on it.
Recurring format: operator playbooks, named data and lead magnets5
AI search visibility as a channel, not a side project1
AI-influenced revenue is undercounted by click attribution1
Channel expansion beyond Meta and Google1
Seasonal concentration and full-funnel spend discipline1
Exit readiness and valuation literacy for founders1
AI search visibility as a channel, not a side project
Howland's most detailed post is a ten-step answer-engine optimisation playbook for DTC brands that have yet to make changes.
Read more
It starts by pulling 20 to 30 real buying questions from support tickets, reviews, site search and post-purchase surveys, using shoppers' own wording.
Then it calls for entering each question word for word into ChatGPT and Google AI Mode and logging whether the brand was mentioned or cited, and which sources the answer drew on.
The remaining steps cover crawler eligibility (GPTBot for training, OAI-SearchBot for search inclusion, ChatGPT-User for user-triggered fetches, with robots.txt changes taking roughly 24 hours to register); keeping price, availability, variants, shipping and returns in visible HTML so the Merchant Center feed matches the page; publishing data only the brand holds, such as return reasons and sizing failures; writing dated comparison pages that name who should not buy; keeping product names consistent across listings and reviews; correcting wrong facts on review sites and forums; deep-linking citations to the relevant product or guide page instead of a homepage; and measuring AI visibility in Search Console and AI referrals in analytics.
Breaks the problem into operator tasks, with a worked example for each stage and sample prompts for discovery, comparison, objection, fit and care questions.
Howland credits the playbook to Lorea Lastiri and says plainly that they are not an AEO expert. They present the framework as someone else's work, adapted for a Meta and Google ads audience.
For brands planning a campaign
For a brand seeking a scoped AI-search engagement, this is a usable starting point: question harvesting, prompt testing, an eligibility audit, a product truth layer, a comparison page and a dated measurement baseline. The depth lies in process design, not proven outcomes, so price a pilot as a bounded build, with the question set and before-and-after citation log as deliverables.
AI-influenced revenue is undercounted by click attribution
Howland summarises Fairing data across 158 DTC brands over 104 weeks, covering brands spending $10M to $2B a year and comparing UTM tracking with post-purchase survey answers.
Read more
Surveys caught 7.5 times more AI-influenced orders than click-based attribution, and the revenue gap was 10.6 times, meaning click methods captured under 10% of the AI-driven revenue surveys found.
Growth was steep: AI's share of gross merchandise value rose 12 times year over year; Adobe tracked AI referrals to US retail sites up 14 times since October 2024; and AI-referred visitors converted 54% better than non-AI traffic, a reversal from the prior year.
Category patterns matter for targeting: high-consideration, high-AOV categories such as mattresses and health devices show the largest AI-driven shares, while in commoditised categories, AI agents default to one or two known brands and smaller brands drop out of consideration.
In housewares, the spread between the most and least AI-attributed brand widened from about two points to more than 20 in a year.
Recommends adding one survey question asking whether an assistant helped the shopper find the brand. Wait for 100+ responses and a 30%+ response rate before acting on the split, then build presence in sources models cite and increase review velocity, citing 17-43% more pricing power when review count doubles.
The figures come from Fairing's tracking and Adobe and Columbia/Yale studies, not Howland's own experimental results.
For brands planning a campaign
Useful for setting client expectations before an AI-search engagement: the value case is a measurement gap, not a traffic spike.
The plan should include a survey change and a review-generation push alongside content work.
The brand-level comparison is a caution for smaller challengers in crowded categories, where losing an agent recommendation can remove them from consideration entirely.
Howland makes the case for AppLovin placements inside mobile games and apps for premium ecommerce, especially brands selling to 40 to 50 year old women, a channel they say they would have dismissed a few years earlier.
Read more
The post explains the two formats: rewarded video, where users are locked in for 15 to 30 seconds to earn in-game currency, and interstitial, where a three-second unskippable window runs before the skip button appears.
The creative advice is placement-specific.
For rewarded video, seconds 0 to 25 tell a story or show a satisfying demo, seconds 26 to 28 build tension, and seconds 29 to 30 carry the strongest call to action, timed for the second before viewers can close it.
For interstitial, the hook should land at second three, when skip becomes tappable—the opposite of Meta, where the hook is expected in the first three seconds.
The cheatsheet: take the three best-performing Meta concepts, recut the timing around the placement interface, run rewarded and interstitial as separate tests, and judge the channel on blended MER.
Recuts existing top-performing creative for the new placement rather than starting from scratch, and tests each format separately.
The performance claims are framed as a 2026-2027 expectation for premium ecommerce, not measured results. Howland notes that in-app attribution is messier than Meta's.
For brands planning a campaign
A concrete brief for testing a second channel: pick three proven Meta concepts, produce rewarded and interstitial cuts that follow the timing rules, run them as separate tests and evaluate on blended MER. Budget for messier in-app attribution; success criteria cannot rely on channel-level ROAS alone.
Seasonal concentration and full-funnel spend discipline
Howland breaks down how HexClad Cookware advertises at low revenue from January to October, while roughly half its revenue lands in November and December, with 12% on Black Friday alone.
Read more
The average order value is $400, rising to $699 for the twelve-piece set.
Fewer than 20% of purchases happen inside a seven-day window, and 80-85% of buyers say they knew the brand for months first.
So October spend runs MER 33% below the year-end target, and the 13 days before the sale run about 50% below it—a deliberate loss the head of growth describes as funding a full funnel.
The ROAS lift peaks above 300% on 12-13 November, before the sale opens, while Black Friday weekend has the lowest lift of the whole period.
The president left about $1M of BFCM budget unspent, on the view that the bottom-of-funnel work was already done.
Offers, timing and targets are locked 15 weeks out, and 85% of November revenue comes from the sale window.
A case study built from named operator quotes and disclosed figures, ending with the two conditions Howland says make it work: faith in the brand and a well-built forecast.
Howland notes that this is not easy to run: brand conviction and forecasting accuracy are prerequisites, not tactics any brand can copy through spend alone.
For brands planning a campaign
Directly relevant to premium brands whose demand is seasonal and considered rather than impulse-driven.
The brief: fund the consideration window in the flat months and resist shifting budget into the peak weekend.
Lock the forecast and offer well before the season rather than optimising during it.
Exit readiness and valuation literacy for founders
Howland offers a DTC valuation calculator to founders, framed by the claim of having helped five DTC brands sell for $500M or more.
Read more
The tool includes a valuation engine that turns a profit and loss statement into adjusted EBITDA and low, base and high multiples; a multiple builder with eight levers, each priced in turns; a bridge from enterprise value to the amount that actually wires, accounting for debt, working capital peg, inventory, fees, rollover and tax; a cost-of-capital normaliser that converts any flat fee or factor rate into a true APR and prices it against nine financing options; and a benchmarks tab with every multiple band and threshold sourced. A worked example takes a $12M brand to $9.7M enterprise value and $5.6M cash at close.
Distributes the tool as a lead magnet through a comment-and-connect request rather than publishing the calculator openly.
The calculator is promoted rather than shown, so readers can evaluate the model's scope and inputs, not the tool itself.
For brands planning a campaign
Signals a founder and owner audience alongside the growth audience: the same people evaluating ad spend are thinking about multiple expansion and clean numbers at exit. For a finance-adjacent or M&A-adjacent partner, that framing is the entry point; a practical engagement would focus on financial modelling and benchmark sourcing rather than media buying.
Recurring format: operator playbooks, named data and lead magnets
Across Howland's posts, they use a consistent shape: a hook with a number or a named brand, then numbered or arrowed steps, then a closing question or resource handoff.
Read more
Depth comes from specifics rather than opinion, including channel mechanics, dollar thresholds, percentages, and direct quotes from the operators being written about, such as the head of growth and president at HexClad.
Third-party studies are cited by name and, in some cases, dated, with Fairing, Adobe, Podscribe and a Columbia/Yale study appearing in the AI attribution post.
They invite comparison from readers, asking whether anyone is seeing AI mentions climb in survey responses, and use two posts to distribute assets, a creative strategy guide and a valuation calculator, through a comment or connection request.
For brands planning a campaign
A brand evaluating this body of work can scope a first engagement around one of the offered assets, with teaching and frameworks grounded in case detail and a resource as the follow-up step. Ask for evidence of delivered client results: the heavy data cited largely comes from other parties' research and named operators' figures.
Curtis Howland says in one post that they run a Meta and Google ads agency exclusively for eight- and nine-figure brands, but the posts never name the agency, so it cannot be listed as a brand.
Read more
No post names a sponsor, employer, investor, or paying client. The named brands are subjects of the posts' analysis or cited data sources, not stated commercial partners; they are listed below as mentions to show where the names come from.
AppLovin
Mention
Post argues that AppLovin's in-app placements, including rewarded video and 3-second interstitials, offer a new arbitrage opportunity for premium ecommerce.
Read more
It gives second-by-second creative timing for both formats and offers Curtis Howland's AppLovin creative strategy guide by direct message. No sponsorship or agency partnership with AppLovin is stated.
Original disclosure
I made an AppLovin creative strategy guide. Send me a connect, comment 'creative' and I'll DM it over.
Case study on how HexClad Cookware advertises ahead of a November-December sales peak, quoting head of growth Connor Rolain and president Jason Panzer on holding back October spend and leaving Black Friday budget unspent.
Read more
The post analyzes the brand; no client or partnership relationship with Curtis Howland is stated.
Post summarizes Fairing's two-year attribution study across 158 DTC brands and says the data will be linked below.
Read more
Fairing is an outside data source for Curtis Howland's argument that surveys catch far more AI-influenced orders than click tracking. No commercial relationship with Fairing is stated.
Start with the long-form operator posts: the ten-step answer-engine optimisation playbook, followed by the post on AI-influenced revenue being undercounted by click tracking.
Together, they show how brands can be found and measured.
Then read the channel and timing pieces—in-game mobile advertising on AppLovin and the HexClad breakdown of funding flat months to win a concentrated November and December—to see how format-specific mechanics carry into paid channels and seasonal planning.
A final post shifts from media buyers to owners, offering a DTC valuation calculator for founders weighing an exit.
Across these posts, Howland uses numbered steps, named third-party data and quotes from the operators being written about, and closes several with an offer to download a resource.
Both offer practical, step-by-step guidance on getting brands cited and recommended inside AI assistants, not just ranked in search results.
Read more
Vijay argues that starting from keywords misdirects AI search work and walks through prompt testing, query fan-outs and the sources models pull. Howland's ten-step playbook approaches the same ground from a DTC operator's perspective, including harvesting real buying questions and logging which sources an answer cites.
Vijay writes for SEO and AI-search practitioners, focusing on reverse-engineering how retrieval works before changing a page. Howland writes for ecommerce operators in a paid-ads agency context, treating the same subject as a channel checklist alongside Meta, Google and in-game advertising, with feed and product-page hygiene rather than ranking theory.
Both audit how brands appear in AI answers and turn their findings into ranked, numbered fixes readers can act on.
Read more
Wright's five-part series reports on roughly 1,800 pages across 11 manufacturing sites and lists issues by how many sites shared them. Howland's AI search post is also a numbered list of concrete tasks, with specific remedies for each.
Wright's posts describe a large-scale technical audit of manufacturer websites, covering schema markup, stale pages, contradictory facts and crawler blocks.
Howland tackles the same problem from the consumer-brand side, focusing on premium DTC categories, buying-question prompts and post-purchase survey measurement.
Wright's other post is career advice rather than channel work.
Both track where ecommerce transactions are moving and what that means for measurement and customer ownership.
Read more
Stephens' post on Best Buy joining TikTok Shop asks who owns the customer relationship and what data a brand gets when a sale happens inside someone else's platform. Howland's AI attribution post makes a parallel argument: click-based tracking cannot see the orders assistants send.
Stephens writes as a platform and product operator, covering social commerce category shifts and interviewing paid-media leaders about campaign-level visibility.
Howland writes as an agency-side adviser, producing executable playbooks and creative timing rules.
The result is a how-to checklist rather than a market read.
Treat this as a placement question first: the audience is a LinkedIn-first professional readership, so the fit turns on whether you want ecommerce operators and growth leads as viewers rather than a consumer or video-first following.
Read more
The only measured audience figure is the profile's LinkedIn audience of 21,724 followers, on 9 October 2026.
The posts do not publish audience demographics, survey results, or a breakdown by country, age, gender, employer, or job function.
The writing consistently targets people running or advising substantial ecommerce businesses: one sets out an AI-search playbook for brands spending eight and nine figures; one presents a 104-week measurement study across brands spending $10M to $2B a year; one compares AppLovin's rewarded and interstitial formats for premium products; one walks through HexClad's seasonal budget discipline using named figures and quoted executives; and one offers a DTC valuation calculator with a worked $12M example.
These topics point to readers making channel, measurement, and budget decisions for high-average-order products.
Follower count says nothing about how far an individual post travels.
The posts provide no per-post engagement data for these five items, so there is no measured basis for estimating reach on a sponsored collaboration.
Ask for platform demographics covering country, age, and gender; audience survey results or another source covering job function and seniority; and recent results for posts on comparable topics, at comparable depth, in the formats you would commission.
These details will help assess whether a premium DTC agency's target buyers are in the audience.
Curtis Howland's posts describe an agency service, not a creator sponsorship program.
Read more
They say the agency runs Meta and Google ads exclusively for high eight- and nine-figure brands.
Two posts offer resources rather than sponsorship packages: an AppLovin creative strategy guide distributed after a LinkedIn connection and a comment of 'creative,' and a DTC valuation calculator introduced with 'Want it?
You know what to do...' These routes lead to their own guides or tools, not verified sponsored-placement menus.
For a brand plan, the practical qualification is revenue scale: the agency service described in the posts is aimed at high eight- and nine-figure ecommerce businesses.
Ask directly about rates, deliverables, usage rights, exclusivity, timelines, onboarding, reporting cadence, contracting terms, and the formal partnership inquiry address.
Paid ads agency engagement
Buyer requirement
The brand should be a high eight- or nine-figure ecommerce business.
Curtis Howland says the agency works exclusively with brands at that revenue level.
Ask directly about scope, fees, minimum spend, and contracting terms.
This profile supports a campaign aimed at a premium ecommerce brand or a DTC growth agency hiring a paid-ads and AI-search operator—not a sponsorship.
Read more
Curtis Howland says they run a Meta and Google ads agency exclusively for high eight- and nine-figure ecommerce brands, putting the hiring client’s revenue band front and centre.
Their posts also outline a usable scope for AI-search work: a ten-step answer-engine playbook that starts by pulling 20 to 30 real buying questions from support tickets, reviews, site search and post-purchase surveys in the shopper’s own wording, then typing each into ChatGPT and Google AI Mode and logging whether the brand was mentioned or cited and which sources the answer pulled.
The playbook covers crawler eligibility (GPTBot for training, OAI-SearchBot for search inclusion, ChatGPT-User for user-triggered fetches, with robots.txt changes taking roughly 24 hours to register); keeping price, availability, variants, shipping and returns in visible HTML so the Merchant Center feed matches the page; publishing data only the brand holds; dated comparison pages; consistent product naming; correcting wrong facts on review sites and forums; deep-linking citations to the right product or guide page; and measuring AI visibility in Search Console and AI referrals in analytics.
For a considered-purchase brand with a seasonal peak, the HexClad breakdown best shows how the work is reasoned: roughly half of revenue in November and December, under 20% of purchases inside a seven-day window, October deliberately run at a marketing efficiency ratio about 33% below target, offers and targets locked 15 weeks out, and about $1M of Black Friday budget left unspent.
The AppLovin post adds a channel-expansion method: take the three best-performing Meta concepts, recut the timing for rewarded video and interstitial placements, test the formats separately, and judge on blended efficiency because in-app attribution is messier.
The posts make a case for hiring: the work is specific rather than opinion-led, the AI-search offer is bounded enough to pilot, and the LinkedIn professional readership of 21,724 followers is a plausible place to reach ecommerce operators.
Before hiring, confirm fees, scope, minimum spend and contracting terms, none of which the posts disclose.
The AEO playbook is credited to Lorea Lastiri, and Howland says plainly they are not an AEO expert; buy it as process delivery with named deliverables.
The heaviest statistics cited (Fairing, Adobe, Podscribe, a Columbia/Yale study) come from other parties’ research, while the brand-level comparison comes from HexClad’s own quoted operators, not client results Howland published.
The audience figure is a follower count, with no audience demographics or per-post engagement data; ask for those before assessing how a collaboration would travel.
When to choose a specialist
A poor fit would be a mass-market consumer brand with a low average order value and a fast, impulse-led purchase cycle, paying for consumer reach through an upfront sponsored video, a TikTok or Instagram activation, or a paid package priced off follower count.
Read more
Three things in the posts point the other way.
First, the only documented commercial format is an agency engagement for high eight- and nine-figure ecommerce brands.
None of the five posts includes a sponsorship menu, rate card, deliverables, usage rights, exclusivity window, timeline or partnership inquiry route, so a creator-sponsorship campaign cannot be qualified from the posts.
Second, the advice is aimed at long, considered purchases and high order values.
The HexClad case assumes a $400 average order value, a two-month selling season, and 80 to 85% of buyers knowing the brand for months beforehand.
The AI attribution post reports the largest AI-driven shares in high-consideration, high-average-order categories such as mattresses and health devices, and warns that in commoditised categories AI agents default to one or two known brands while smaller brands drop out of consideration entirely.
A low-AOV impulse product sits at the opposite end of both.
Third, the measurement method requires volume and time a short consumer campaign will not have: the recommendation is to wait for 100 or more survey responses at a 30% or better response rate before reading the AI discovery split.
Reach also cannot be estimated for a consumer push: the only measured audience figure is a LinkedIn following of 21,724, with no audience demographics, audience survey or per-post engagement data, and no video, newsletter, interview or long-form transcript to show how they perform in consumer-facing formats.
A brand that needs a low-funnel consumer placement with documented sponsorship terms should look elsewhere, or first ask for the audience and format information needed to assess the campaign.
A worked campaign example
Illustrative campaign: a premium home-goods brand with an average order value around $300, revenue in the tens of millions, a demand peak concentrated in roughly two months of the year, and a growth team that already runs Meta and Google but cannot tell whether AI assistants are driving sales.
Scope a 12-week paid-ads and AI-search engagement rather than a long retainer: the strongest posts show process design, and the brand should be able to see whether the process produces anything before committing further.
Weeks 1 to 2: harvest 20 to 30 buying questions from support tickets, reviews, site search and post-purchase surveys, keeping the shopper’s own wording, and split them across discovery, comparison, objection, fit and care.
Add the single AI discovery question to the post-purchase survey at the same time, and agree in advance not to read the split until there are 100 or more responses at a 30% or better response rate.
Type each question word for word into ChatGPT and Google AI Mode more than once, logging whether the brand was mentioned or cited and which sources the answers pulled, with the date on every entry so the log can be re-run later as a before-and-after comparison.
Weeks 2 to 5: audit eligibility by checking which OpenAI crawlers are currently allowed and deciding whether to block the training crawler; note that robots.txt changes take about 24 hours to register.
Build the product truth layer: price, availability, variants, shipping and returns in visible HTML, with the Merchant Center feed matching the page, and consistent product, model and variant names across listings, old reviews and current SKUs.
Publish data only the brand holds—return reasons, sizing failures, product tests—which the playbook calls the most underused asset in DTC.
Correct wrong facts on the review sites and forum threads surfaced in the question log.
Weeks 5 to 10: publish one dated comparison page naming the criteria and who should not buy, and deep-link citations and ads to the relevant product or guide page rather than the homepage.
Run the paid creative test in parallel: take the three best-performing Meta concepts, recut them for rewarded video (story or demo through second 25, tension through second 28, strongest call to action in the final seconds) and interstitial (hook landing at second three, when the skip button becomes tappable), run the two formats as separate tests, and judge them on blended marketing efficiency ratio because in-app attribution is messier than Meta’s.
Weeks 10 to 12: read out AI visibility in Search Console, AI referrals in analytics, and citation rate on the same dated question set, then set the next quarter’s scope from what moved.
Write these deliverables into the agreement: the dated question set and citation log, a crawl-eligibility note, the published comparison and guide pages, the survey change with its threshold, and a test readout covering both the AI visibility work and the channel test.
Before signing, confirm fees, scope, minimum spend and contracting terms, which the posts do not disclose.
The AEO playbook is credited to Lorea Lastiri, and Howland says it is not their own expertise; name who delivers the AI-search work and ask what information about prior client outcomes is available beyond the posts.
Choose Aabhas Vijay when a DTC growth agency's bottleneck is AI-search strategy: whether ChatGPT searches the web for the brand's category, how query fan-out works, and which sources and competitors get cited.
Read more
Their posts give an agency a diagnostic sequence to run before hiring anyone: pick 5 to 10 real buyer questions, run them in ChatGPT, check whether live retrieval happens, look for fan-out searches (the post cites an average of 7.6 per prompt), and map the sources the model repeatedly finds.
A second post explains why keyword-led AEO fails: optimizing for the shopper's question rather than the AI's research process, producing content around topics instead of decisions, missing the retrieval opportunity, and confusing keyword ranking with AI visibility.
For a premium ecommerce brand with an unresolved AI-visibility problem, that diagnostic work is more specific than Curtis Howland's ten-step playbook.
The tradeoff is that Aabhas Vijay's posts do not say they run paid Meta or Google ads for high eight- or nine-figure ecommerce brands, or discuss DTC seasonal planning, AppLovin creative tests, or a product-feed and creative loop.
The same campaign decision still needs a paid-ads operator, and the posts do not say they provide one.
Ask for the exact scope of the consultation offer and evidence of client results; neither is confirmed by the saved post text.
Choose Grace Wright when the buyer needs an AI-visibility audit across a large website estate, especially for a considered-purchase or manufacturing business where the issue is whether assistants can find, trust, and quote the company—not media buying.
Read more
Wright's post reports an audit of about 1,800 pages across 11 manufacturing websites, with findings ranked by frequency: missing or generic schema on 11 of 11, thin third-party corroboration on 11 of 11, pages that cannot yield a clean answer on 11 of 11, stale content on 10 of 11, contradictory facts on 9 of 11, orphaned pages and broken proof on 9 of 11, proof locked in PDFs on 8 of 11, crawlers blocked or throttled on 8 of 11, missing or JavaScript-injected H1s on 3 of 11, and live placeholder content on 2 of 11. For a premium ecommerce brand with a sprawling catalogue or legacy site, that audit method is a better first step than Curtis Howland's playbook when the root cause is technical or related to content credibility.
The tradeoff is category fit.
The posts describe an audit of manufacturing websites, not DTC ecommerce, and do not cover paid Meta or Google management, AppLovin placements, seasonal demand planning, or AI-attribution measurement.
The campaign needs both paid-ads and AI-search expertise; Grace Wright's posts address the AI-search diagnostic half, not the paid-media half.
A second post offers career-habit advice from an intern-to-director path, rather than campaign information for DTC ecommerce.
Choose Kait Stephens when the decision shifts from hiring a paid-ads and AI-search operator to understanding social commerce and platform-owned transactions.
Read more
Stephens's posts analyse Best Buy's arrival on TikTok Shop with nearly 10,000 products, the difference between beauty and higher-consideration categories, TikTok driving discovery and checkout while Best Buy handles fulfilment and returns, and the customer-ownership problem when a sale closes inside someone else's platform.
The second post shows a brand-side digital commerce operator bringing a paid-media leader onto an advisory board for an ad platform product.
It also names the measurement problem: top-line ROI without campaign-level visibility, and last-click attribution described as not worth making business decisions on.
For a premium ecommerce brand weighing TikTok Shop or seeking an operator's view on paid-media measurement, that is closer to the decision than Curtis Howland's agency and AEO process posts.
The tradeoff is that the posts offer strategic commentary and an advisory-board announcement, not a documented paid-ads agency engagement, AEO playbook, or DTC seasonal plan.
They do not describe crawler eligibility steps, a question-set method, product truth layer, comparison-page process, or AppLovin creative recut.
If the need is an operator to run Meta or Google ads and AI search for premium ecommerce, Kait Stephens offers platform and social-commerce context, but not the hands-on delivery information Curtis Howland's posts provide.
Ask for commercial terms and evidence of client outcomes; neither is disclosed.
This profile draws on the full text of five LinkedIn posts by Curtis Howland, covering answer-engine optimisation, AI attribution measurement, in-game mobile advertising, a seasonal campaign breakdown of HexClad Cookware, and a DTC valuation calculator.
The posts do not say when they were published; the linked profile account was recorded on 9 October 2026.
The figures and resources the posts cite — Fairing's attribution research, Adobe and Podscribe numbers, a Columbia/Yale study, an AppLovin creative guide and the valuation calculator — are described by Howland, but the posts do not say whether they were independently opened.
No books, shows, newsletters, videos or interviews were available, so a search for this name may surface work that is not covered here; that remains an open research gap rather than a settled answer.
