In partnership with
It’s Wednesday, a perfect day to waste some marketing money!
Your media plan includes a line called retargeting, and it’s been getting quietly worse for three years. This week somebody said why.
Automated agents and bots now account for more than half of all web requests, per Cloudflare, and Digiday spent Monday asking agency buyers what that does to a campaign. The answers were unusually plain. GoFish's clients have moved "out of programmatic and back into places where people are more comfortable like Google, Amazon, Meta." At Collective Measures, bot traffic was "the final nail in the coffin" for one client's retargeting. LERMA/ says retargeting performance "has steadily declined." RPA sees bots producing fraudulent purchases and scrambled attribution. Then the twist: not every brand wants the machines out.
Mellow Sleep, a pillow brand, wants the good agents in: "We're less interested in blocking automated traffic than in classifying it." John Lewis is cited the same way. The visitor that reads your product page might be a scraper, a fraud bot, or a customer's shopping assistant, and the same firewall rule treats all three the same.
So the retargeting pool you keep paying to refill is part people, part bots, and part agents that might buy. A filter cannot tell them apart. First-party data and a sorting rule can.
Today’s sponsor:
Join CX leaders from Clay, Anthropic, and more at Pioneer
AI is fundamentally transforming CX, making perfect experiences possible. But there's no playbook for delivering them.
Join Pioneer on October 7th to explore how CX leaders are redefining what's possible with AI.
You’ll learn how some of the most innovative leaders from Clay, Anthropic, Gamma, and more are transforming their organizations and reinventing their operations.
Plus, you’ll hear from Fin product leaders on their vision for Fin and be the first to see what’s new.
TALK ON THE MARKET
Alex Lieberman, co-founder of Morning Brew, on Thursday, after lunch with a strategy executive at a 5,000-person company:
The bulldozer, he says, is the head of AI, who owns the strategy and its execution.
Deciding which bots to let in is the kind of question that dies without both. Too technical for the boss, too commercial for IT, so it lands on the media agency, which decides by moving your money somewhere the question does not come up.
NUMBERS MATTER
$7.1 billion. Clay's valuation after a $115 million round led by Wellington, Reuters reported on Wednesday, more than double a year ago. Its agents analyze business data and act on it for sales teams. Customers include Google and Anthropic. CEO Kareem Amin: "Now we're building agents that can help grow your company for you."
OUR PARTNER: AIJOBLIST
Every wondered how AI models are trained?
How about becoming an AI trainer yourself (and making money as one)?
DAILY PROMPT
You are a data analytics expert specializing in marketing performance monitoring. Your task is to analyze marketing data and identify anomalies that warrant investigation.
## Data Context
I'm providing marketing performance data from [TIME_PERIOD]. The baseline metrics are [BASELINE_METRICS]. Recent performance shows [CURRENT_METRICS].
Key channels/campaigns: [LIST_CHANNELS_CAMPAIGNS]
Historical average performance: [HISTORICAL_BASELINE]
## Anomaly Detection Framework
Analyze the data using these dimensions:
1. **Statistical Anomalies**: Identify metrics that deviate significantly (>15-20%) from historical averages or expected ranges. Flag both positive and negative deviations.
2. **Temporal Patterns**: Look for unusual timing patterns—spikes/drops that don't align with typical weekly/monthly cycles, campaign schedules, or seasonal trends.
3. **Cross-Channel Correlations**: Identify unexpected relationships between channels. For example, if one channel typically drives traffic to another, flag when this correlation breaks.
4. **Cohort-Level Changes**: Detect anomalies in specific audience segments, geographies, devices, or demographics that differ from overall trends.
5. **Attribution Anomalies**: Flag unusual changes in conversion paths, first-touch vs. last-touch attribution shifts, or unexpected channel contribution changes.
## Output Structure
For each anomaly identified, provide:
- **Anomaly**: Clear description of what changed
- **Severity**: High/Medium/Low based on business impact
- **Confidence**: Your confidence level (High/Medium/Low) based on data quality
- **Likely Causes**: 2-3 potential explanations (technical issues, competitive activity, platform changes, campaign adjustments, external events)
- **Recommended Actions**: Specific investigation steps or immediate responses
- **Metrics to Monitor**: What to track going forward to validate the anomaly
## Investigation Priorities
Rank anomalies by:
1. Business impact (revenue/conversion effect)
2. Urgency (requires immediate action vs. monitoring)
3. Actionability (can we influence it)
Provide a prioritized list with the top 3-5 anomalies requiring immediate attention.Use this before the next budget review, because the platform will not run it for you. Digiday's buyers found the problem by looking. The dashboards had been reporting reach all along.
CMO CORNER
A well-designed system should run without applause or panic.
Bot traffic is applause you did not earn. Sessions go up, reach climbs, everyone relaxes. Then the conversion rate says something is wrong and the panic starts. Both come from the same mistake: reading activity as if it were people.
Build the reporting so it counts what a human did, and only that. A smaller number, a calmer room.
We provide consulting services for CMOs on AI strategy for marketing teams through our Teamless offering.
TOOL UPGRADE
First-party data is king, so own the channel:
GetResponse — best for mid-size teams that want email, automation and landing pages in one place. Try GetResponse here →
Systeme.io — best for lean teams replacing five tools with funnels, email and checkout under one login. Try Systeme.io here →
Nuelink — best for pushing what you already own, blog, podcast, catalogue, out to social without a person doing it. Try Nuelink here →
→ We have 300+ curated, filtered, and categorized AI tools in our directory. Browse to find the best ones that help your marketing team.
LEARN THIS
"First-party data is king" is easy to say and expensive to mean. The AI Data Strategy Framework is the part in between: what to collect, where it enters, who owns each field, and what an agent may do with it.
For the leader whose retargeting just died and who would rather rebuild on data than on someone else's audience.
→ We publish workshops, courses, guides, manuals, and more on AI marketing for free. Browse our learning resources for free.
BRAND SLOP
A political campaign is a brand with a sixty-day runway, and this one put a synthetic version of its competitor on air.
Texas Senate nominee Ken Paxton released an ad using an AI-generated James Talarico to deliver "some of his most controversial statements," Axios reported, the real quotes combined and repackaged into a performance that never happened.
Nate Persily of the Stanford Law AI Initiative: "The more artificial media that we consume, the less likely that people will be able to discern what is true from what is false, and also the less that they might care."
That last clause is the one to keep. Not that people will be fooled. That they will stop minding. Comparative advertising has always wanted to put words in the competitor's mouth. Now it can use their mouth. Decide what your brand will not do with that before someone proposes it as a bold idea.
The brief 330K+ marketers actually read
TLDR Marketing is the free daily brief that 330K+ growth marketers, performance marketers, and CMOs actually read. The most interesting stories in marketing, curated and summarized in 5 minutes.