Mickaël Aubard shares practical analysis of performance marketing, including cross-channel ad benchmarks, campaign strategy, and tools that help teams manage advertising across platforms.
Cross-channel CTR post names sources and notes Snapchat click definition changed reported clicks.
Lower-funnel-only Snapchat spend fades because you run out of people.
Google Performance Max may source videos from advertiser pages or social accounts by default.
Check first: Confirm sponsorship offer, rates, audience analytics, and whether claims are independently verified.
Main ideas
A benchmark is only useful if you name the sample and check the definition · Lower-funnel-only spend fades because you run out of people · Read the platform defaults before you trust them
Mickaël Aubard shares practical analysis of performance marketing, drawing on examples from Google, Meta, Snapchat and other channels.
His posts cover ad benchmarks, campaign strategy, platform settings and tools for managing advertising across platforms.
A cross-channel CTR post, for example, pairs figures with named sources and notes that a change in Snapchat’s click definition affected reported clicks.
The reviewed material does not establish whether Aubard has signature books, shows, businesses or earlier work that also explain searches for his name; that identity research remains to be done.
Aubard’s work connects marketing advice to concrete operational choices.
He explains why lower-funnel-only Snapchat campaigns can lose reach, lays out a proposed full-funnel budget split, and gives examples of tactics and reported results.
In a separate post, he flags that Google Performance Max may source videos from an advertiser’s pages or social accounts, then gives a specific setting to check.
Across these examples, he pairs practical next steps with figures or mechanics, making the posts useful to readers evaluating campaign decisions.
The corpus also includes product-adoption and partner posts, so it is not exclusively independent channel analysis; claims and benchmarks are presented in the posts and are not independently verified here.
Across seven LinkedIn posts, Mickaël Aubard works as a performance-marketing practitioner who turns platform mechanics into decisions a team can act on: what a normal click-through rate looks like on each channel, how much budget belongs above the lower funnel, which default settings carry hidden risk, and how to bring AI into routine media buying without losing human checkpoints.
Read more
The recurring habits are sourcing numbers rather than asserting them, checking the measurement definition before comparing, reading defaults and usage rights before scaling, and treating in-person partner conversations as a form of research. Where the material relays someone else's framework or third-party research, the post names the person or the data source.
A benchmark is only useful if you name the sample and check the definition
Aubard treats channel benchmarks as reference points that need context, not headline numbers.
Read more
The CTR post gives a median for each channel alongside the sample size or data provider, and it flags where a measurement change makes a comparison misleading. The practical point for a reader is that a platform's number can shift because the platform changed what it counts, not because performance moved.
The post lists Google Search at 6.64% median from 13,474 US campaigns, Meta ecommerce at 2.19% from 35,000 brands via Triple Whale, TikTok at 1.77% overall but 0.5% to 0.7% for ecommerce, email at 1.69% for campaigns and 5.58% for automated flows via Klaviyo, LinkedIn at 0.44% to 0.65% for sponsored content and 2.68% for Thought Leader Ads, YouTube in-stream B2B at 0.51%, and display at 0.22%, or 0.68% for dynamic formats. For Snapchat it gives 0.94% and notes reported clicks fell 7.56% in January when the click definition changed.
Lower-funnel-only spend fades because you run out of people
Aubard's takeaway from a Snap office talk by Lars van Egmond is that the problem is not the channel but the audience: starting with lower funnel only works until the same known people have been retargeted repeatedly.
Read more
Van Egmond's reframe, relayed in the post, is that full funnel is not brand versus performance but brand fuelling performance, because incremental reach comes from the algorithm going after new people.
Aubard passes on a starting budget split and a peak-season checklist built on that idea.
The proof points are attributed to outside research and the talk, not to Aubard's own controlled tests.
The post reports that the best-performing Snap advertisers put almost 25% of budget into upper and mid funnel (Fospha), and describes a Lulus example where the same budget spread across the funnel reached 90% more unique users, cut CPM by 52% and produced slightly higher ROAS.
The suggested split is 25% upper funnel optimised for impressions, 75% lower funnel, with 10% of the lower-funnel budget on retargeting.
Season tactics include starting 30 days before Black Friday, adding upper and mid funnel to always-on for +50% ROAS, and multi-format DPA for 28% better CPA.
Aubard's method on platform automation is to explain what the feature actually does, say where it is switched on, and then give a concrete action.
Read more
The argument is that a default setting can expose assets a brand or creator never intended to run as paid creative, so the check belongs before scaling, not after.
The Performance Max post explains that Google finds videos on a landing page, YouTube channel and socials, adds up to 10 per asset group, swaps in new ones on a recurring basis, and re-uploads them to a Google-owned YouTube channel with the advertiser's logo.
It notes the setting is on for every new PMax campaign, that some existing campaigns may have it on too, and that Google's page says nothing about usage rights, so an organic-only creator video could run as a paid ad.
Aubard recommends turning off Sourced videos in every PMax campaign under Asset optimization in asset group settings, then checking Manage sources before turning it back on.
AI pays off when it encodes your own rules into repeatable commands
Rather than treating AI as a general writing tool, Aubard frames it as a way to compress recurring media-buying routines into named commands that already know the account's targets and rules.
Read more
The distinction he draws is between an adequate media buyer and an elite one: the elite version keeps human approval on budget moves and re-checks the numbers before anything goes to a client.
The post lists 11 Claude commands, including /kill-or-keep for scale, cut, refresh or leave-alone calls per ad; /morning-read for overnight campaign movement; /budget-step to raise winning ad sets 15% at a time with approval; /lost-share for Google campaigns capped by budget; /copy-five to write five angles from one product page; /hook-to-page to check landing headline alignment; /client-note for weekly updates with checked numbers; /recheck-numbers to recompute CPA and ROAS before sending; /interview-first to make Claude ask questions before building; /dump-to-brief to turn a voice note into a one-page brief; and /sound-human to strip AI tells from client-facing text.
The blocker to testing a new channel is rebuild time, not the creative
Aubard repeatedly argues that teams already suspect their winning ads would travel to another platform, and that what stops them is the manual work of rebuilding campaigns.
Read more
The fix he describes is import automation that maps objectives, structure, targeting and budgets across, so the test can happen in minutes. The adoption numbers he cites are activity signals rather than performance comparisons.
The Snap Paris post describes importing a Meta campaign into Snapchat in a few clicks, with objectives, ad set structure, targeting and budgets mapped automatically, creatives, copy and headlines carried over and editable in bulk, and lead forms already mapped. The Datashake post shows the same friction from the agency side: 49 team members on the AdManage account, 57 people in a shared Slack channel, more than 71,000 ads launched and 11,481 in October alone, a first rollout that followed a Black Friday of uploading every ad by hand, and an internal channel for sharing time-saving tips.
In-person partner conversations are a working research method
Several posts treat a coffee or an office visit as the point of the trip rather than a courtesy.
Read more
Aubard uses those meetings to pick up what agencies and platform teams are seeing, and says plainly that sitting across the table beats a video call. This is the least metric-driven of the recurring habits in the material, but it explains how the other posts get their field examples.
Posts cover coffee with Antoine Laborie from ASight by the Canal Saint-Martin, described as one of the most impressive agencies Aubard has come across; a visit to the Datashake office in Paris; and a meeting at Snap Inc.'s Paris office with Morgan Azogui, where the discussion covered events the teams are planning together and where the industry is heading as more teams go multi-platform.
Answers the newcomer question: what click-through rate is normal on each major channel, and where do those numbers come from?
Read more
Mickaël Aubard collects median CTR benchmarks with named sources: Google Search 6.64% from 13,474 US campaigns; Meta ecommerce 2.19% from 35,000 brands via Triple Whale; TikTok 1.77% overall but 0.5% to 0.7% for ecommerce; email 1.69% for campaigns and 5.58% for automated flows via Klaviyo; Snapchat 0.94%, with reported clicks falling 7.56% in January when the click definition changed; LinkedIn 0.44% to 0.65% for sponsored content, 0.02% for text ads and 2.68% for Thought Leader Ads; YouTube in-stream B2B 0.51%; and display 0.22%, or 0.68% for dynamic formats. The post lists the source for each benchmark so a reader can check the context.
For brands planning a campaign
Use this as a starting benchmark list for paid ads planning, but match channel, format and measurement definitions before comparing a client's CTR. The Snapchat click-definition note is a warning that year-over-year CTR comparisons can reflect measurement changes rather than performance changes.
Explains why a lower-funnel-only strategy fades and how a full-funnel budget split is meant to fix it, with the argument and the numbers in one place.
Read more
Mickaël Aubard reports a Snap office talk by Lars van Egmond.
The pattern described: launch lower funnel only because ROAS works, then it fades as the same known audience is retargeted.
The reframe is that full funnel is not brand versus performance; it is brand fuelling performance.
Evidence includes almost 25% of budget from the best-performing Snap advertisers going to upper and mid funnel via Fospha; a Lulus example with the same budget spread across the funnel reaching 90% more unique users, 52% lower CPM and slightly higher ROAS; plus channel-level ROAS claims from TransUnion and Triple Whale.
The starting split is 25% upper funnel optimised for impressions, 75% lower funnel, with 10% of the lower-funnel budget on retargeting.
Peak-season tactics include starting 30 days before Black Friday, adding upper and mid funnel to always-on for +50% ROAS, multi-format DPA for 28% better CPA, creator content in Sponsored Snaps for 30% higher purchase rate and 6x higher ROAS, and First Snap takeovers for 42% higher ad awareness and 27% higher ROAS.
For brands planning a campaign
Useful for a Snapchat or retail peak-season plan where a brand has plateaued with retargeting. Test incremental reach and upper-funnel budget before assuming the channel has stopped working, and treat the ROAS figures as claims to validate with a controlled test.
Answers an urgent operational question: which videos can Performance Max use, and what should a brand or creator check before trusting default settings?
Read more
Mickaël Aubard explains that Performance Max finds videos on a landing page, YouTube channel and socials, adds up to 10 per asset group, swaps in new ones on a recurring basis, and re-uploads them to a Google-owned YouTube channel with the advertiser's logo.
It is on for every new PMax campaign, and some existing campaigns may have it on too.
Google's page says nothing about usage rights, so an organic-only creator video could run as a paid ad.
Aubard recommends turning off Sourced videos in every PMax campaign under Asset optimization in asset group settings, then checking Manage sources before turning it back on.
For brands planning a campaign
A direct pre-campaign check for brands and creators: audit PMax asset settings and source lists before launch. For creator partnership work, clarify paid usage rights for organic video, because the default setting can turn organic content into paid creative.
Shows a practical way to use an AI assistant for recurring ad-account tasks, with one-word commands rather than abstract prompts.
Read more
Mickaël Aubard lists 11 Claude commands for media buying.
Account tasks include /kill-or-keep for scale, cut, refresh or leave alone per ad; /morning-read for overnight campaign movement; /budget-step to raise winning ad sets 15% at a time with approval; /lost-share for Google campaigns capped by budget; /copy-five to write five angles from one product page; /hook-to-page to check landing headline and hook alignment; and /client-note for weekly updates with checked numbers.
Chat tasks include /recheck-numbers to recompute CPA and ROAS before sending, /interview-first to make Claude ask questions before building, /dump-to-brief to turn a voice note into a one-page brief, and /sound-human to strip AI tells from client-facing text.
Each skill is set up to know targets and rules.
For brands planning a campaign
A model for an internal ad-ops assistant: turn repeated checks and reporting into named commands, then keep human approval on budget changes and client sends. Useful for teams evaluating AI workflow productivity or building a training resource.
Gives a concrete adoption story for an ad-management tool, including team usage, a manual-upload trigger and a partner feedback loop.
Read more
Mickaël Aubard reports on a Paris agency visit.
Datashake has 49 team members on the AdManage account and 57 people in a shared Slack channel.
More than 71,000 ads have launched through AdManage, including 11,481 in October, described as the agency's biggest month.
Estelle leads the October leaderboard with 1,289 ads launched.
The team has an internal channel for sharing AdManage time-saving tips.
The story began nearly a year earlier after a Black Friday manual-upload problem, and the first testers were Hugo, Elisa, Clara and Marie-Emmanuelle.
For brands planning a campaign
For an ad tech vendor, this is a case-study pattern: show team-level adoption, a painful trigger and a customer feedback loop rather than only product features. A brand or agency can use it to judge whether an ad-management tool fits high-volume creative operations.
Explains the friction that stops proven Meta campaigns from being tested on Snapchat, and shows what an import automation does.
Read more
Mickaël Aubard describes a partner meeting at Snap's Paris office.
The pattern described: brands know their winning Meta ads would work on Snapchat, but rebuilding everything by hand never reaches the top of the to-do list.
The automation imports a Meta campaign into Snapchat in a few clicks; objectives, ad set structure, targeting and budgets map automatically; creatives, copy and headlines carry over and are editable in bulk; lead campaigns move across with lead forms already mapped.
Aubard frames the result as testing proven winners on a new channel in minutes instead of days.
For brands planning a campaign
Useful when a brand is stuck at the first step of a multi-platform test: the bottleneck may be campaign rebuild time rather than creative or budget. The post is also a product-adoption story for an ad tech vendor, but the cited speed claim is from the post, not an independent benchmark.
In-person partner and agency meetings as a research format3
Meta-to-Snap campaign portability and agency adoption2
Cross-channel ad benchmarks and what good looks like1
Full-funnel Snapchat strategy and peak-season planning1
Platform automation risk: PMax sourced videos1
AI-assisted media buying workflows1
Cross-channel ad benchmarks and what good looks like
Aubard publishes platform-by-platform CTR benchmarks for 2026, covering Google Search, Meta ecommerce, TikTok, email, Snapchat, LinkedIn, YouTube in-stream B2B, and display.
Read more
The post gives median figures with sample sizes or source names, such as 13,474 US Google Search campaigns, 35,000 brands from Triple Whale, and Klaviyo email data. Aubard also notes that Snapchat's reported click figure fell after a click-definition change, which matters when comparing channels.
Publishes medians with sample sizes and source names, then points readers to a fuller guide for checking their own results.
The Snapchat figure includes a click-definition change caveat, so the number is not directly comparable to earlier periods.
For brands planning a campaign
Useful for a paid ads benchmark report because the material connects channel-level metrics to their sources and flags definition changes that can make comparisons misleading. A brand can evaluate Aubard's cross-channel metric literacy from the specific numbers and caveats rather than from a general claim of expertise.
Full-funnel Snapchat strategy and peak-season planning
Aubard argues that lower-funnel-only campaigns fade because they keep targeting people who already know the brand.
Read more
The post reframes full funnel as brand fuelling performance and shares a starting split of 25% upper funnel, 75% lower funnel, with 10% of the lower-funnel budget on retargeting. It also reports case and partner figures, including Lulus reaching 90% more unique users and 52% lower CPM at the same budget, plus a peak-season playbook that says advertisers who started 30 days before Black Friday saw 5x higher ROAS last year.
Uses a partner talk and third-party measurement points to support a budget-allocation framework, then gives a concrete starting split and a peak-season checklist.
The proof points are attributed to outside research and a partner talk, not to Aubard's own controlled testing.
For brands planning a campaign
A brand can evaluate Aubard's ability to translate full-funnel theory into budget splits and seasonal timing for Snapchat. This is directly relevant to campaign planning, especially for brands deciding how much spend to place above the lower funnel and when to start Black Friday activity.
Aubard explains that Google Performance Max can now run videos a brand never uploaded, and that the sourced-videos setting is on by default.
Read more
The post says Google finds videos on the landing page, YouTube channel, and socials, adds up to 10 per asset group, swaps in new ones on a recurring basis, and re-uploads them to a Google-owned YouTube channel with the brand's logo. Aubard highlights that Google's page says nothing about usage rights, so an organic-only creator video could run as a paid ad, and recommends turning off sourced videos in every PMax campaign.
Walks through how the feature works, where the setting lives under Asset optimization, and what to check before turning it back on.
For brands planning a campaign
Important for paid media governance and creator-content rights. A brand can evaluate Aubard's practical risk analysis, and the post gives a specific setting to check before scaling campaigns that might pull in creator assets.
Aubard writes that Claude can make someone a decent media buyer, but lists 11 skills that make a media buyer elite.
Read more
The post presents each as a one-word command, including /kill-or-keep for scale, cut, refresh, or leave alone decisions; /morning-read for overnight campaign changes; /budget-step for 15% ad-set raises; /lost-share for capped Google campaigns; /copy-five for new angles; /hook-to-page for landing-page alignment; /recheck-numbers for CPA and ROAS checks; and /sound-human for removing AI tells from client-facing writing.
Turns media-buying routines into named commands that already know targets and rules, so one word starts the job.
For brands planning a campaign
A brand can evaluate Aubard's AI adoption framing for performance teams. The post is useful for a benchmark or training report that needs to show how AI commands can be tied to everyday media-buying checks rather than general AI commentary.
Meta-to-Snap campaign portability and agency adoption
Aubard describes AdManage users activating Snapchat alongside Meta and importing winning Meta campaigns into Snapchat in a few clicks.
Read more
The post says objectives, ad set structure, targeting, budgets, creatives, copy, headlines, and lead forms map automatically, so teams can test proven winners on a new channel in minutes instead of days. A separate post reports Datashake as a heavy AdManage user with 49 team members on the account, 57 people in a shared Slack channel, more than 71,000 ads launched, and 11,481 ads in October alone.
Combines a partner meeting with agency-office metrics and customer stories to show product usage, adoption scale, and internal team behavior.
The Datashake figures are adoption and activity numbers, not a controlled performance comparison against other platforms.
For brands planning a campaign
A brand can assess how Aubard connects product automation to multi-platform expansion. The material is relevant for teams considering Snapchat alongside Meta because it focuses on reducing rebuild friction and on agency-level adoption signals.
In-person partner and agency meetings as a research format
Several posts are framed around in-person visits and partner conversations, including coffee with Antoine Laborie from ASight, a visit to the Datashake office in Paris, and a meeting at Snap Inc.'s Paris office with Morgan Azogui.
Read more
Aubard uses these meetings to discuss marketing, SaaS, multi-platform direction, and how brands are getting started on Snapchat. The format is relationship-led field research rather than a written benchmark or tutorial.
For brands planning a campaign
This shows a format a brand can evaluate for partner-led content, event coverage, or market-pulse posts. It is less directly tied to cross-channel ad metrics than the benchmark and workflow posts, but it signals an approach built on direct industry contact and ongoing partner relationships.
Aubard's retained posts show a small set of business ties: he speaks as a promoter or representative of AdManage, describes a working partnership with Snap, and documents Datashake as a heavy AdManage user.
Read more
A separate post records an agency meeting with ASight. The posts do not state whether Aubard owns or is employed by AdManage, so that relationship is recorded as an unstated disclosed partnership rather than ownership or employment.
AdManage
Disclosed partnership
Aubard speaks as a builder or representative of AdManage in posts about importing Meta campaigns into Snapchat and about agencies running ads through the platform.
Read more
The posts promote AdManage's automation and report usage, but do not state whether Aubard owns or is employed by the business. The posts say "we automated it" and refer to AdManage users without naming his role.
Aubard describes a Snap partner meeting at Snap Inc.'s Paris office, discussing AdManage users activating Snapchat alongside Meta, planning events together, and working with the Snap crew weekly.
Read more
He also shared takeaways from a Snap office talk. The posts thank Snap staff and promote Snap campaign imports; they do not state a contract or payment.
Aubard describes Datashake as one of the agencies with the most people on AdManage: 49 team members on the account and 57 people in a shared Slack channel.
Read more
He reports more than 71,000 ads launched through AdManage and 11,481 in October alone, and describes visiting the Paris office. The post shows a substantial client relationship to the AdManage account; it does not state whether Datashake contracts with Aubard personally.
Mickaël Aubard writes a running field notebook on performance marketing.
Read more
Start with the 2026 cross-channel CTR benchmark post to see how Aubard handles numbers: each channel gets a median, a sample source and a caveat where the definition changed.
From there, the Snapchat full-funnel post gives a budget split and peak-season checklist drawn from a partner talk, the Performance Max sourced-videos post shows how Aubard reads platform defaults and usage rights, and the Claude command post shows how recurring media-buying checks become named commands.
The Paris office-visit posts show a second, relationship-led way Aubard collects examples from agencies and platform teams.
The supplied material is a small set of recent LinkedIn posts, so use it to sample current themes rather than as a full-career archive.
Kazim's posts also work the performance-marketing problem of scaling spend across channels.
Read more
The retained posts describe moving a DTC brand from $200K to $1.2M a month by adding TikTok, YouTube, Snapchat, AppLovin and CTV, and frame Q4 growth as a diversification question rather than a Meta-and-Google-only question. Like Aubard, Kazim gives campaign structure and sequencing that a reader can compare.
Kazim writes from an agency scaling-narrative perspective, with client case studies and webinar-style lead generation, while Aubard's posts stay closer to platform mechanics, benchmarks and tooling or automation.
Morgan's posts share Aubard's interest in why performance degrades as spend rises.
Read more
Morgan connects Meta CPM and ROAS pressure to landing pages and funnel segmentation, and argues that a scaling ceiling often sits in the site rather than the ad algorithm. Both writers treat full-funnel and creative-to-page alignment as core media-buying questions.
Morgan focuses on the ecommerce store and landing-page experience for seven- and eight-figure Shopify brands; Aubard ranges more widely across channel benchmarks, ad platform defaults, AI workflows and cross-platform campaign operations.
Kraatz's posts are also tactical ecommerce ad guidance.
Read more
The retained posts cover Black Friday and Cyber Monday Meta campaign mistakes and testimonial ad formats, including when to retarget, when not to test new creative, and how social proof can be scaled from static tests. Like Aubard, Kraatz gives concrete campaign moves rather than general marketing commentary.
Kraatz's material is concentrated on Meta ad creative and seasonal playbooks for ecommerce brands; Aubard covers a broader set of channels and platform tools, with more emphasis on measurement, default settings and automation.
For a paid ads benchmark report, the useful question is whether Mickaël Aubard's LinkedIn following is made up of people who plan, buy, or measure paid media.
Read more
The only measured audience evidence supplied is a follower count: 10,014 followers on the public LinkedIn profile, observed 8 October 2026.
That number shows the size of the audience reachable on one channel; it does not show who is in it.
The post content gives indirect signals about appeal.
The material is written for people working inside ad accounts: cross-channel CTR benchmarks by platform, Google Performance Max asset sourcing, Snap full-funnel budget splits, bulk ad-launch workflows, and one-word prompts for media-buyer tasks.
The posts read as working notes shared with peers rather than brand announcements, and they name agencies, platforms, and tool teams around Mickaël Aubard (Datashake, Snap, ASight, AdManage).
A following that responds to that material is likely to skew toward agency and in-house performance marketers rather than a general marketing readership, but that is an inference from the subject matter, not a measurement of follower composition.
Not covered by the evidence: geography, seniority, job function, employer type, and age or gender composition.
No audience survey, no platform analytics export, and no per-post reach or engagement figures were supplied, so none are estimated here.
For a buyer, the concrete next step is to ask Mickaël Aubard for recent account analytics or an audience survey covering follower geography, seniority mix, and industry, plus reach and engagement on posts about comparable topics and formats.
If that cannot be supplied, treat the match between following and benchmark-report readers as unverified and test any placement on a small scale rather than assuming a broad performance-marketing audience.
LinkedIn followers
10,014
Public LinkedIn profile for Mickaël Aubard · 2026-10-08
The available material contains no sponsorship offer, media kit, rate card, or paid-partnership listing for Mickaël Aubard.
Read more
The partnership activity that does appear is company-level and platform-facing: Aubard posts about AdManage, the ad-management tool whose users and features they describe, and about its working relationships with Snap, the Paris agency Datashake, and ASight.
Those posts cover automated Meta-to-Snapchat campaign import built for AdManage users, events being planned with Snap, and agencies running thousands of campaigns through the platform.
For a buyer planning spend, the practical consequences are concrete.
There is no published format menu, deliverable specification, or rate to build a plan around, and no audience breakdown tied to a sponsored placement.
The benchmark and platform content Aubard publishes does name third-party data sources — Fospha, Triple Whale, TransUnion, Klaviyo, Google, Meta, TikTok, LinkedIn — but citing a source is not a sponsorship, and a brand appearing in that content would not by itself indicate a paid arrangement.
Any paid partnership would need to be negotiated directly, and the most useful basis for a proposal is the work already visible: cross-channel ad benchmarks, platform and tool analysis, and hands-on campaign tactics across Meta, Snapchat, Google, TikTok and LinkedIn.
Brief: a B2B ad tech company is commissioning a paid ads benchmark report for 2026 and wants a named performance-marketing voice to front it — someone who can present cross-channel numbers, explain where each figure comes from, and show media buyers how to use them.
Read more
This is a strong fit on the evidence reviewed.
Aubard's published work is built around a per-channel CTR benchmark set with named sources and sample sizes: Google Search 6.64% across 13,474 US campaigns, Meta ecommerce 2.19% from 35,000 brands via Triple Whale, TikTok 1.77% overall but 0.5% to 0.7% for ecommerce, email 1.69% for campaigns and 5.58% for automated flows via Klaviyo, Snapchat 0.94%, LinkedIn 0.44% to 0.65% for sponsored content and 2.68% for Thought Leader Ads, YouTube in-stream B2B 0.51%, and display 0.22%.
That post also flags that Snapchat's reported clicks fell 7.56% in January when the click definition changed — exactly the kind of measurement caveat a benchmark report needs and most sponsored content skips.
Around the numbers, the same material shows operational judgment a report audience will recognise: the Snapchat full-funnel budget split of 25% upper funnel, 75% lower funnel with 10% of that on retargeting, sourced from a partner talk and third-party measurement; and the Performance Max warning that sourced videos are on by default, with no statement on usage rights, so an organic-only creator video could run as a paid ad.
The content is written for people inside ad accounts, and the audience evidence supplied is a public LinkedIn following of 10,014 as observed on 8 October 2026, with the composition unmeasured.
Two limits to plan around.
Audience makeup — geography, seniority, job function — is not covered by any supplied evidence, so ask for recent account analytics or an audience survey and treat the match as unverified until then.
And there is no documented sponsorship offer, media kit, or rate card, so a paid arrangement would be negotiated directly rather than pulled from a published menu.
When to choose a specialist
Brief: a consumer DTC brand wants a paid creator placement — a sponsored short-form video or lifestyle review for a household product, with paid reuse rights and a reach-based performance guarantee.
Read more
This does not fit the reviewed material, for three concrete reasons.
First, no sponsorship offer is documented.
Across the reviewed posts there is no format menu, deliverable list, rate, or inquiry page, so there is nothing to build a consumer placement around.
That is a gap in the evidence, not proof that sponsorships are never accepted; it means a buyer would be negotiating blind.
Second, the audience can't be matched to the brief.
The only measured figure is 10,014 LinkedIn followers, and nothing in the supplied evidence describes geography, seniority, or buying role.
Consumer reach on short-form video channels is not evidenced at all.
Third, the format and content type don't line up.
The reviewed material is ad-platform operations analysis — cross-channel CTR medians, Snapchat budget splits, Performance Max asset settings, bulk campaign workflows, one-word AI commands for media buying — plus partner and agency visits.
It is written for practitioners running ad accounts, not for consumers choosing a product.
No video or transcript evidence was supplied, so there is no basis for judging on-camera performance.
One practical point cuts against a reuse-rights brief in particular: Aubard has publicly warned that Performance Max can pick up videos from landing pages, YouTube channels, and socials and re-upload them as paid creative, with no usage-rights statement from the platform.
A brand that intends to run creator video as paid media should settle rights in writing regardless of who it works with.
Here, that would mean starting from scratch.
A worked campaign example
Worked angle for the benchmark-report brief: "The 2026 Cross-Channel CTR Reality Check."
The spine is the eight-channel comparison Aubard already publishes, kept in the order a media planner reads it: search, Meta, TikTok, social, email, LinkedIn, YouTube, display.
Each figure stays attached to its source and sample size in the body of the report, not in a footnote — 13,474 US Google Search campaigns, Triple Whale's 35,000 brands, Klaviyo's email data, TransUnion and Triple Whale for the Snapchat ROAS context.
The reason is that the report's value is being checkable, and a planner who can trace a median to its sample will defend it in an internal review.
Three added modules carry the report beyond a table.
One, a definition-change section built on the Snapchat example: reported clicks fell 7.56% in January after the click definition changed, which is the clearest available illustration that a year-over-year CTR drop can be measurement, not performance.
Two, a budget-structure section showing the Snapchat starting split of 25% upper funnel, 75% lower funnel with 10% of the lower funnel on retargeting, presented as a tested starting point with its third-party evidence rather than a universal rule.
Three, a governance checklist covering the Performance Max sourced-video setting — where it lives under Asset optimization, what Google's documentation does not say about usage rights, and why a brand should audit source lists before launch.
Sequencing: publish the benchmark table first as the lead asset, follow with the definition-change piece because it is the most shareable and the most useful to anyone who reported a CTR decline, then the budget-structure and governance modules as a two-part practitioner series. Package the media-buying command set — /kill-or-keep, /morning-read, /recheck-numbers and the rest — as a downloadable checklist tied to the report, since those commands map directly onto the weekly routines of the people the report is for.
What to measure: whether readers can find the source of any figure from the page, whether they use the checklist in a live account, and whether the definition-change module changes how they present a CTR decline internally.
What to agree before starting: audience analytics covering geography and seniority, the sourcing rules for any figure not already published, and commercial terms, since no sponsorship offer is on record and the arrangement would be negotiated directly.
What this angle does not solve: audience composition on any channel other than the LinkedIn follower count, and consumer-facing reach, neither of which the reviewed material covers.
Choose Kazim when the benchmark report needs a documented scaling case study rather than a median table.
Read more
The reviewed posts walk a DTC health and wellness brand from $200K to $1.2M per month with the moves broken out by month: TikTok launched alongside YouTube and contributing 18% of revenue at a lower CPA than Meta within 30 days; creative volume moved from 8 to 35 new creatives a week through a modular hooks-bodies-CTAs system; audience architecture split into cold, warm and hot layers with separate creative and bid strategies; then Snapchat, AppLovin and CTV added, each assigned a funnel role. A brief that wants a practitioner narrative of how spend was added across surfaces without losing signal fits that material more directly than an eight-channel CTR comparison.
The tradeoff is checkability.
Those figures read as one agency's client result; the reviewed posts name no external data source or sample size for them, and the second reviewed post is a live-session promotion, so no published method sits behind the numbers.
Aubard's benchmark set attaches a named source and sample size to each channel median — Klaviyo for email, Triple Whale for Meta.
A reader who wants to trace a figure to an audited sample gets less from Kazim; a reader who wants a step-by-step scaling story gets more.
Choose Morgan when the report turns on platform cost pressure and funnel diagnosis rather than cross-channel CTR medians.
Read more
The reviewed material tracks Meta CPMs from $9.84 to $18.01 over two years, ROAS from 2.57x to 1.56x, and CAC up 31% to $52.86, then argues that targeting moved into creative and the landing page, and that most brands still send their most expensive visitors to a generic page.
A second post covers the store-as-anchor problem for 7- and 8-figure Shopify brands and message-market fit drifting away from new audiences as spend scales.
A report section on why efficiency is eroding and what to change structurally fits that more closely than a channel benchmark table.
The tradeoff is sourcing and scope.
Morgan cites Common Thread Collective and The DTC Index for the platform figures, and both reviewed posts promote a webinar series rather than publish the underlying method.
The material is Meta- and site-side, so a reader who needs comparable search, email, LinkedIn and display medians still would not get them.
Aubard's eight-channel set covers that breadth; Morgan goes deeper on why the Meta numbers are moving against brands.
Choose Kraatz when the deliverable is a seasonal Meta playbook rather than a measurement report.
Read more
The reviewed posts give a BFCM mistakes list with what to do instead — start campaigns early and cite Particl's finding that 28% of 5,000 DTC companies launched by 24 October, skip new creative testing in favour of updating proven ads, and shift budget to retargeting — plus an argument for testimonial ads that starts with static ads on a small budget to find which reviews convert, then scales into new formats. A brief that wants platform-specific seasonal execution detail fits that material more directly.
The tradeoff is cross-channel coverage and date currency.
Kraatz's material is Meta-focused; search, email, YouTube, LinkedIn and display medians are absent, and the reviewed posts date to November 2024 and February 2025, before the click-definition and Performance Max changes the 2026 benchmark work addresses.
A reader choosing between the two for a cross-channel benchmark report loses the current multi-channel picture but gains ready-made seasonal tactics they can act on immediately.
This profile draws on seven LinkedIn posts by Mickaël Aubard: a 2026 cross-channel click-through-rate benchmark list for eight channels, a Snapchat full-funnel budget framework relayed from a partner talk, Google Performance Max sourced-video settings, a set of Claude commands for routine media buying, a Meta-to-Snapchat campaign import, an agency adoption story, and Paris office visits with agency and platform teams.
The supplied text does not state exact publication dates; the posts refer to 2026 benchmarks, October campaign activity and Black Friday planning.
This is a small recent sample rather than a full-career archive, so no long-term trend is claimed, and third-party figures are reported as the posts present them.
No books, shows, businesses or earlier-work evidence was supplied, so searches that depend on that part of Aubard's identity are not yet covered here.
Recent engagement
Across their last 6 LinkedIn posts, the median post got 22 reactions and 4 comments. Checked 2026-11-02.
