Source credit: This commentary summarizes and applies HubSpot’s 2026 State of Marketing Report, created in collaboration with Datalily. The introduction in the supplied report is by Kipp Bodnar, HubSpot’s chief marketing officer. Recommendations specific to Metropulse are editorial analysis, not statements or endorsements by HubSpot.
AI has made content easier to produce, but it has not made trust, originality, or customer understanding easier to earn. HubSpot’s central argument is that successful brands will combine AI’s efficiency with human judgment, a recognizable point of view, and a better understanding of their audiences. Its Loop Marketing framework describes this as four connected activities: express your story, tailor it to your audience, amplify its reach, and evolve through continuous learning.
For Metropulse, that suggests a shift from asking, “How much traffic can we attract?” to asking, “Which visitors value our work enough to return, identify themselves, and eventually pay?”
The supplied HubSpot report supports investigating that shift. In its survey of 1,505 marketers, 58% said AI referral traffic had higher intent than traditional search traffic, while 93% reported that personalization improved leads or purchases. These are marketers’ reported experiences—not a prediction of Metropulse’s conversion rate. They suggest an opportunity to learn more about readers who arrive, rather than assuming that more visits will automatically produce more revenue.
Metropulse must also distinguish between an AI crawler and a person arriving through an AI recommendation. The crawler may collect content; the referred reader may become a customer. Even among crawlers, discovery and model training can be separate functions. OpenAI, for example, allows publishers to permit its search crawler while disallowing its training crawler.
With no advertising income and commission arrangements that have not worked, MP has reason to question the value of indiscriminate traffic. But blocking automation is a cost-control and content-protection decision—not, by itself, a monetization strategy. A subscriber business requires a second proposition: something readers consistently find useful enough to support.
The opportunity is not simply to put existing content behind a signup form. It is to discover what readers cannot easily get from a generic AI answer, then build a dependable offering around that difference: original reporting, distinctive interpretation, carefully verified curation, timely alerts, practical resources, or access to informed people.
End of proposed summary.
1. Filter selectively—not indiscriminately
My recommendation is to separate traffic into four working categories before imposing broad restrictions:
| Traffic category | Proposed MP treatment | Commercial reasoning |
|---|---|---|
| Abusive scrapers, spam, and excessive automated requests | Block or rate-limit based on evidence | Protect resources and reduce abuse |
| Verified search and discovery crawlers | Generally preserve access to selected public pages | Maintain opportunities for readers to discover MP |
| AI training crawlers | Make an explicit publisher-policy decision | Decide whether access provides sufficient value to MP |
| Human readers, including AI-referred visitors | Give them a clear path to relevant content, signup, or purchase | Test actual customer potential |
Selective treatment is technically possible: OpenAI separates search and training crawler controls, while Cloudflare documents rules that distinguish verified bots from other automated traffic. Verification matters; a crawler’s claimed name is not enough on its own.
Do not equate a short visit, foreign location, or failure to subscribe with “garbage.” Those signals alone do not establish that a visitor is a bot or commercially worthless.
Pros and cons of the proposed shift
| Decision | Potential benefit | Main risk or limitation |
|---|---|---|
| Remove abusive automation | Cleaner operational data and less resource consumption | Overbroad rules can interfere with useful access |
| Preserve useful discovery | Keep a route from search and AI answers to MP | Discovery does not guarantee referrals or revenue |
| Build a free email audience | Establish a permission-based channel for learning and repeat contact | A free list costs time and does not prove willingness to pay |
| Offer paid membership | Test direct reader revenue rather than depending on advertisers or commissions | Requires a recurring benefit and continued delivery |
| Gate most content immediately | Make payment expectations clear | Ask readers to commit before they understand MP’s value |
For MP, I would test a hybrid: public discovery content, a narrowly focused free briefing, and one small paid offering. That follows HubSpot’s emphasis on distinct positioning, audience-specific messaging, and rapid experiments without requiring MP to adopt its software or every channel in its framework.
2. Determine what to sell before building a large subscriber program
The missing fact is not simply “Who visits?” It is “What valuable job do readers need MP to do?”
Use audience behavior and interviews together. HubSpot’s customer-behavior guidance recommends segmenting audiences, identifying each group’s key benefit, combining quantitative and qualitative evidence, and then testing changes against conversion and acquisition metrics.
A practical MP research sequence
- Identify the strongest content clusters.
Look for repeat visits, meaningful article engagement, relevant clicks, replies, and completed signups—not just pageviews. - Interview a small initial group of engaged readers.
Start with approximately 10–15 conversations as an exploratory exercise, not a statistically representative survey. - Ask about actual behavior.
“What did you come here to accomplish?” and “What do you currently pay for to solve that problem?” are more useful than “Would you support MP?” - Propose one specific paid benefit.
State what it includes, how often it arrives, who it serves, and its price. - Seek an actual purchase or a clearly disclosed paid pilot.
Interest and survey answers are preliminary evidence. Payment tests the commercial proposition more directly.
Candidate offerings to investigate
These are hypotheses, not claims about MP’s current audience:
| If readers primarily need… | Test this offering | Evidence to seek |
|---|---|---|
| Less time spent finding trustworthy information | A concise, human-edited briefing | Repeat use and paid renewals |
| Help making a particular decision | A one-time guide, comparison, or research package | Purchases tied to a specific need |
| Ongoing interpretation of a changing topic | A specialist subscription | Continued engagement and retention |
| Access to informed discussion | A moderated briefing or member Q&A | Participation and repeat attendance |
| A way to sustain MP’s public work | A supporter membership | Explicit willingness to fund the mission |
Do not force a subscription onto an occasional need. If readers want one answer once, a one-time purchase may be the better initial test.
The failed commission arrangements also deserve a short review: Did readers click? Did merchants record purchases? Was attribution reliable? Were the offers relevant? Were commissions too small? Failure at one stage does not establish that readers will never buy anything.
3. Structure the subscriber link around a benefit
Avoid making “Subscribe” the entire pitch. Tell readers what they receive.
Suggested navigation and article links
- Main navigation: “Get the MP Brief”
- Article-specific invitation: “Want more reporting on this topic? Get the MP Brief.”
- Separate paid link, once an offer exists: “Explore MP Membership”
Use “free briefing” and “paid membership” consistently so readers do not confuse email signup with a purchase.
Suggested free-signup copy
Get the MP Brief
Human-edited reporting and perspective on [specific subject], delivered [actual frequency].
Understand what matters—and why—without sorting through the noise.Email address: [field]
Button: “Send me the free MP Brief”By signing up, you agree to receive the MP Brief by email. Unsubscribe anytime. See our privacy notice.
Replace the brackets before publishing. Promise only a frequency and editorial standard MP can reliably maintain.
Keep the initial form short. My proposed setup is:
- Required: email address.
- Optional: one topic preference, only if MP will use it.
- After signup: email confirmation and an optional questionnaire.
- Separate choice: permission to participate in research interviews.
Use signup abuse controls and email confirmation, but do not treat a confirmed address as proof of a human reader or likely purchaser.
4. Put the questionnaire after signup
The questionnaire should help MP select an offering—not collect personal information without a defined use. Audience research is most useful when it connects reader needs with observed behavior and subsequent tests.
Suggested questionnaire
| Question | Suggested response format |
|---|---|
| 1. What brought you to MP today? | Search engine / AI assistant / social media / recommendation / already a reader / other |
| 2. Which topic would you most like MP to cover? | Select up to two from MP’s actual editorial categories |
| 3. What are you trying to understand or accomplish? | Short optional text |
| 4. What is hardest about getting this information elsewhere? | Finding trustworthy sources / too much information / missing context / insufficient detail / other |
| 5. Which MP offering would be most useful? | Briefing / original reporting / practical guide / alerts / expert Q&A / none |
| 6. What, if anything, do you already pay for that meets this need? | Optional text; include “nothing” |
| 7. May we contact you for a brief reader interview? | Separate yes/no choice |
After you have a concrete paid product, test its price separately:
MP is considering [specific deliverable], delivered [frequency], for [price].
Would you: join a paid pilot / need more information / prefer a one-time purchase / not buy this?
A “join” response should lead to an accurately described checkout or enrollment process—not be counted as revenue until payment occurs.
5. Measure whether the remaining audience supports a business
Use separate reporting for automated requests, estimated human readership, free subscribers, and paying customers. Do not divide paid subscriptions by all server requests and call that a meaningful reader conversion rate.
My proposed weekly scorecard:
| Metric | What it helps answer |
|---|---|
| Estimated human visits by source | Where potential readers come from |
| Confirmed free signups ÷ estimated human visits | Whether the signup proposition attracts readers |
| Paid customers ÷ people shown the paid offer | Whether the specific offer sells |
| Paid conversion by topic and source | Which audience-offer combinations merit attention |
| Repeat purchases or renewals | Whether value persists beyond the initial sale |
| Refunds, cancellations, and reasons | Where the promise fails to match experience |
| Revenue less delivery, payment, and acquisition costs | Whether growth improves the economics |
Include the cost of editorial labor and support. With no ad revenue, “more subscribers” can still mean larger losses if the offering costs more to deliver than it earns.
A simple planning equation is:
Break-even paying members=Monthly price per member−Variable cost per memberMonthly fixed costs
Use actual retention data before projecting long-term subscriber value.
6. A staged 90-day test
Days 1–30: Establish the baseline
- Review traffic and bot classifications before broad blocking.
- Preserve a rollback path for security-rule changes.
- Identify the strongest reader needs and content clusters.
- Conduct initial reader interviews.
- Launch one clear free-signup proposition.
Days 31–60: Test one paid promise
- Choose one audience and one deliverable.
- Publish a sample that demonstrates the value.
- Offer a limited paid pilot with clear terms.
- Record purchases, objections, fulfillment time, and costs.
Days 61–90: Decide what deserves expansion
- Look for continued use, repeat payment, and manageable delivery costs.
- Refine the offer if readers value it but resist its packaging or price.
- Stop or redesign it if payment and continued use do not materialize.
- Expand only after the economics—not merely the signup count—justify it.
This applies HubSpot’s loop to MP at a manageable scale: articulate a distinctive promise, learn which readers it serves, distribute it where they discover information, and improve it using actual results. The report is useful strategic context, but it is not evidence that MP’s remaining traffic will sustain a subscription business. That requires MP’s own paid experiment.