David Herrmann

In their view, automation-first work treats consumers as data points for a model to imitate rather than people worth connecting with.

• They say clients hire a marketer's brain, gut, and experience; automate all of that, and the work stops mattering.
• They are not against using AI for tasks but object to it becoming a marketer's entire personality and focus.

They use the film They Live, where sunglasses reveal that mass media is being used to control people, to frame what they see across Slack communities, LinkedIn, DTC Twitter, and marketing podcasts. They argue that the conversation focuses on AI doing more, not on good marketing, and urge marketers to keep pushing on what good means.

Read original

They say one to three ads per ad set is a recipe for burnout, while running more than ten creatives in a set has sustained scaled budget and success for longer than a month.

• They do not specify a sweet spot.
• They recommend flexible ad format when creative looks and feels similar and targets the same type of person, and standard format when creative looks and feels different but targets the same person.

The post is a numbered set of working notes.

• The first point says ad sets with one to three ads burn out, while higher creative volume—above ten—has held up longer.
• The second sets a rule for choosing flexible versus standard formats based on how similar the creative is.

Read original

They say to optimize for 9x16 vertical first and abandon 4x5 and 1x1 statics.

• If using a static, add video motion for Reels placements.
• They also emphasize upkeep: periodically update engaged and existing customer lists inside Meta, coordinate with the email and retention team on those lists, use UTMs on site links, keep the ad-level optimize setting on, and point ads to best sellers, top colors, and accessories rather than one generic destination.

The same six-point post lists five practical notes: use vertical-first creative, add motion to statics for Reels, update lists alongside email and retention teams, use UTMs, and keep the ad-level optimize toggle on so ads can highlight specific products and options.

Read original

What they say is usually left out is how to use the audiences.

• They recommend creating custom audiences from the engaged brand and creator Instagram accounts, naming them so they are not confused with brand audiences, and targeting similar, like-minded creators with them as a mid-funnel play.
• They note that this works especially well over the holidays.

The post describes building custom audiences from engaged pages and Instagram accounts, labeling them clearly, and targeting them toward similar creators as a solid middle-of-funnel play that performs well during the holiday period.

Read original

They joke that they must be old and out of touch because they are not on current best-of lists or asked to be on podcasts, then cite a childhood Book It pizza reading achievement as the only recognition they ever cared about. The post is a personality and community aside, not operational marketing advice.

They say the only popularity list that ever mattered to them was the Book It pizza list in first through fourth grade, which they kept on their resume, and that none of the younger DTC crowd can claim that. The joke signals familiarity with the community, skepticism about its popularity cycles, and a willingness to make themselves the punchline.

Read original

If you open one item, make it this one.

A numbered post of six practical notes on running Meta ads as a DTC brand.

• The points, in order: ad sets with only one to three ads burn out, and David Herrmann reports that running more than ten creatives per ad set has sustained both scaled budget and performance for longer than a month, though they name no exact sweet spot.
• Flexible ad format suits creatives that look and feel alike and address the same kind of person; standard format suits creatives that look and feel different while still targeting the same person.
• Whitelisting still works.
• The step they say is usually left out is building custom audiences from the engaged brand and creator Instagram accounts, naming them clearly to distinguish them from brand audiences, then targeting those audiences against similar creators as a mid-funnel play—a tactic they note works well over the holidays.
• Build creative for 9x16 vertical first, abandon 4x5 and 1x1 statics, and add some video motion to statics for Reels placements.
• Update engaged and existing customer lists inside Meta periodically in step with the email and retention team.
• Add UTMs to site links, leave the ad-level optimize setting on, and point ads to best sellers, top colors, and accessories rather than one generic destination.

For brands planning a campaign

For a Meta audit, turn these six points into a checklist: count creatives per ad set; review the format and whether it fits how similar the creative actually is; check that engaged-page custom audiences exist and are clearly named; confirm 9x16-first asset production; confirm list hygiene with the retention team; and verify UTMs and the optimize toggle. Start creative testing with the ad-set density claim: test batches of more than ten creatives per ad set against the brand’s current one-to-three structure, varying volume rather than creative alone.

Read it for the stance, not the tactics.

David Herrmann borrows the plot of the film They Live, in which sunglasses reveal that mass media is being used to control people, and applies it to what they see across Slack communities, LinkedIn, DTC Twitter, and marketing podcasts: nearly all the conversation is about using AI for marketing rather than about marketing itself.

• They argue that automation-first marketing treats consumers as data points for a model to imitate instead of people to connect with; clients hire a marketer’s brain, gut, and experience, and automating all of that makes the work worth little.
• The post makes clear that AI is fine to use for things; the objection is to making it a marketer’s entire identity and focus.

For brands planning a campaign

This signals how David Herrmann would approach an account: human-led creative judgment with AI as a tool, not AI-generated volume as the strategy.

• If a brief assumes large-scale AI-assisted content production, raise that assumption early.
• The post states a position, not a performance claim; ask for evidence before treating it as proof they have tested AI-driven campaigns.

It teaches no tactics.

David Herrmann jokes that they must be old and out of touch with DTC because they are not on current best-of lists or podcast invites.

• They then pivot to a childhood Book It pizza reading list as the only recognition they ever cared about, which they say they still keep on their resume.
• It is a self-deprecating joke about visibility in the community.

For brands planning a campaign

Useful mainly as a read on voice and on how they position themselves relative to the DTC scene; it is not evidence of audience reach, partnerships, current demand, or results. Do not build a campaign argument on this post.

They say 1–3 ads per ad set burns out, while higher creative volume—more than 10 in a set—has sustained scaled budget and success for longer than a month.

• They recommend flexible ad format when creative looks and feels the same and speaks to the same type of person, standard format when creative differs, and prioritizing 9x16 vertical assets over 4x5 and 1x1 statics.
• The post also advises adding motion to statics for Reels, updating engaged and existing customer lists, staying aligned with email and retention teams, and using UTMs and the ad-level optimize toggle to highlight best sellers, colors, and accessories.

Audit method: compare current ad set volume and formats, then test higher-volume sets, vertical-first creative, list updates, and UTM/optimization settings.

For brands planning a campaign

A DTC brand assessing paid social should ask David Herrmann to audit ad set structure and creative volume, test more distinct assets per set, shift production toward vertical-first video, and review audience-list hygiene and UTM/optimization settings.

Examples: Source 1

It emphasizes creating custom audiences from engaged pages and Instagram accounts, using naming that distinguishes them from brand audiences, and targeting similar, like-minded creators with those audiences as a mid-funnel play, especially over the holidays.

Whitelisting plus custom audiences built from engaged pages/accounts, aimed at similar creators for mid-funnel reach.

For brands planning a campaign

For a DTC brief, include a whitelisting and creator-adjacent audience test in the Meta audit. Check whether the brand segments engaged social audiences separately from customer lists and whether holiday mid-funnel plans use them.

Examples: Source 1

They argue that automation and slop treat consumers as data points rather than people to connect with, and say clients hire a marketer's brain, gut, and experience. The post supports using AI but warns against making it an entire personality or letting it dull human idea generation.

For brands planning a campaign

For campaigns involving AI-assisted creative or content at scale, David Herrmann is likely to push for human strategy, original ideas, and authentic consumer connection. Treat that as a counterweight in an AI-heavy DTC media plan.

Examples: Source 1

It shows a self-deprecating, community-aware voice and familiarity with DTC social circles.

This is a personality and community post, not a method or performance claim.

For brands planning a campaign

Consider the tone when matching a creator to a brand voice; the post supports a self-aware, humorous register.

Examples: Source 1

Rego's framing of a stuck DTC engine — too few ads launching, nothing hitting the scale line, money in the wrong place, or traffic that will not convert — overlaps directly with the argument that 1-3 ads per ad set burns out and creative volume is the lever.

Rego's posts are diagnostic and financial: plotting every ad by lifetime spend against CPA, drawing kill and scale lines, then calculating marginal ROAS from a geo-holdout test and mapping a cash-payback waterfall month by month. Herrmann's writing stays closer to day-to-day account setup and format rules, without the measurement modeling or cash-flow arithmetic.

Read original · Original 2

Luby writes that growth strategy was once equivalent to Meta strategy and now splits into feed strategy and market strategy, and that a brand's relationship with the algorithm is a lease someone has to manage.

• Luby also points to how heavily short-form vertical video dominates the creative spread of seasonal DTC brands.
• That is the same ground Herrmann covers when arguing against a pre-2025 Meta playbook and for vertical-first assets.

Luby writes about the marketing leadership job — naming and describing the Chief Marketer role, how attention is earned on platforms a brand does not control, and how honestly to size a market — and describes scaling a drinkware brand from low six figures to eight figures. Herrmann's posts are narrower: the mechanics of creatives, formats, audiences and settings inside one ad account.

Read original · Original 2

Arif's account of learning that by losing money on their own stores, and of starting from customer reviews, comments, objections and buying triggers before opening a design tool, sits behind the same push for creative that has a job to do.

Arif writes first-person career and process narrative — design background, an unsuccessful store, the founding of a creative studio, and how a brief gets filled in — and describes running a creative partner for brands spending $500K+ a month. Herrmann's posts offer account structure, format and audience advice, plus an argument about AI.

Read original · Original 2

Their posts lay out a diagnostic method: plot every ad ever run with $100+ spend on a log-spend versus CPA graph, shade the testing zone under $1,000, add target CPA, a kill line at 1.5x target, and a scale line at 0.7x, then read the dot pattern to distinguish a creative volume problem, a creative quality problem, a media buying problem, and a conversion problem.

• The companion post tackles a cash question growing DTC brands often miss: add a post-purchase survey asking when and where customers first heard about the brand, build a payback curve, run a geo-holdout to estimate marginal ROAS on the next dollar rather than average ROAS, and model the cash trough and payback month at different spend levels, with a worked example of $100K/month added to a $1M base.
• Rego is the stronger choice when finance sign-off and payback timing are the real constraints on the campaign.

The method assumes data the brand may not have: post-purchase survey responses by channel, enough geo coverage for a holdout, and clean revenue attribution.

• It stays at the level of graph, bottleneck, and cash model, rather than the day-to-day account mechanics David Herrmann covers, such as creative count per ad set, flexible versus standard format, engaged-page custom audiences, or aspect ratio defaults.
• A brand adopting Rego's framing still needs someone to decide what goes into the ad sets.

Examples: Source 1 · Source 2

Their Chief Marketer post frames the role as Feed Strategy (earning attention on platforms the brand does not control through content, distribution, community, and trend spotting) plus TAM Strategy (sizing the market honestly and pushing product and distribution toward a bigger one).

• It argues that in 2026 growth strategy means both together, not Meta alone.
• The second post makes the video case concretely by looking at DTC seasonal brands before the 4th of July and reporting roughly 1,000 active Meta creatives with about 90% video.
• It concludes that short-form vertical video is the language of the internet and a marketer in 2026 needs to speak it fluently.
• Luby is the stronger choice when a brand is deciding whether to restructure how the marketing function works and shift the production slate decisively to video.

This is observation and framing, not an audit procedure.

• It gives the brand a reason to move budget toward video and a vocabulary for the marketing role, but no account-level checklist: ad-set creative counts, format decision rules, custom audience setup, or UTM and list-hygiene steps.
• The competitive read is a snapshot of other brands' creative spread, not a controlled result, so it points to a direction rather than a test design.
• Ask for an account-level checklist if the brand needs implementation guidance.

Examples: Source 1 · Source 2

Their posts describe a performance creative shop, Wasl Creative, working as a partner for DTC brands spending $500K+ a month across paid socials and delivering 3,000+ statics and emails a month with a 90% creative approval rate across 20-30 active accounts.

• The second post explains the working method: they learned performance creative by running their own stores and losing money, and concluded that a designer's instinct—does this look good?—is the wrong question.
• Before opening Figma, they dig into the customer through reviews, comments, competitors, objections, and buying triggers, because the best creative ideas sit in the customer's words rather than a generic design request.
• Arif is the stronger choice when the brand needs a repeatable pipeline of statics and email creative built from customer language, with volume and approval-rate detail to inform a creative testing plan.

The posts cover creative production and concepting, not media buying.

• They provide no ad-set structure, budget allocation, kill and scale rule, or payback model, so a brand still needs the account audit and cash view from the other options.
• The performance figures describe the agency's own output, not client outcomes: the posts give no client names, spend levels, or before-and-after CPA movement.
• Ask for client results if those are needed to assess campaign performance.

Examples: Source 1 · Source 2