SODP Dispatch - 24 September 2026

How to know if you can trust an AI’s answer to your question

 By Leo S. Lo

I recently typed a simple question into Google search: How much screen time is too much for teenagers? Instead of presenting links, as Google had been doing for many years, it gave me an AI-generated answer. The artificial intelligence agent cited a number, then complicated that reply, noting that quality and balance of time could matter more than the number of hours, and that “too much” time could depend on a teenager’s sleep, exercise, school demands and mood.

I tried another search: Should I take a daily aspirin? This time the AI answer presented me with medical information, warned about risks and offered more tailored guidance if I provided my age and medical history.

These were good replies. What interested me was that they were different kinds of replies.

Debate about AI answers has focused on accuracy: Did the system get the answer right? That matters, but accuracy is only one test. Each kind of answer requires a user to judge something different.

I find it useful to sort AI answers into an “answer typography” of four broad types: factual, interpretive, constructive and strategic. A factual claim can often be checked against a source. An interpretation can be accurate and still reflect choices about which evidence matters. A construction can be well reasoned and still be wrong for the person receiving it. A beautifully written strategic document may not be true. Yet AI presents all four types of answers in much the same fluent, authoritative form; the differences are easy to miss.

I’m university librarian and dean of libraries at the University of Virginia who leads national efforts to develop AI competencies for library professionals, and I consult widely on AI literacy. I first proposed the typography in the Journal of Academic Librarianship.

The four categories are not airtight boxes. A response from an AI agent can reflect several types. That said, I describe each type of answer below, and offer guidance for deciding whether a reply is ready to use or needs more investigation.

💬 We want to hear from you: Next time an AI gives you an answer, do you stop to ask what type it is (factual, interpretive, constructive, or strategic) before trusting it? Hit reply and tell us; we read every response, and yours might shape a future issue.

📊 PubTech 2026: Build, Buy, Own with WordPress VIP

The plumbing of the web is being redone. Some publishers have already re-evaluated their architecture for it; others are still running on the same stack as 2019. That gap is the theme of this year’s conference, answering three questions: what's worth building, what you're still paying for that you'd never buy today, and what you own that someone should be paying you for.

Forty publisher executives spend a day answering these questions about their own businesses. It's the 4th annual edition, running as one hybrid day, in Central London and online worldwide, and everyone leaves with the room's answers in writing.

This is for publishing leaders who decide what gets built, what gets cut, and what your business charges for. If you are a CEO, Managing Director, Founder, CPTO, CTO, CPO, or Chief Digital Officer, this room is built for you.

Date: Tuesday, November 10, 2026
Time: 09:00 – 21:00 London (US edition 15:30 London / 10:30 ET)
Venue: The UnHerd Club, 6 Old Queen St, London SW1H 9HP, United Kingdom

What you will walk away with:

• The Ownership Audit — a one-page snapshot of your own build-buy-own position across CMS, data, paywall, ad stack and audience

• The Ownership Ledger — the room's answers on Build, Buy and Own, consolidated, anonymised and circulated to everyone who attended

• A benchmark of where your build-versus-buy line sits against 40 peers

• Three bets for 2027, committed to by the room and sent to you afterwards as a post-event ebook

Confirmed speakers for 2026 – so far

• Charlotte Cijffers — Chief Digital Officer, ITP Media Group

• Pete Barr-Watson — Chief Product & Technology Officer, The Independent

• James Hill — Publishing Lead EMEA, WordPress VIP

• Mads Holmen — Head of Product, Platform, MonetizationOS

• Courtney Jarrett — Head of Product, Customer Experience, MonetizationOS

• James Henderson — Founder & CEO, MonetizationOS

The day is high-trust and entirely off-the-record under Chatham House Rule. No panels, no product demos, no slideware, just working sessions built around what publishers own, what they're still paying for, and what's worth building themselves.

Seats are strictly limited to 40, two per business, and every application is answered personally within two working days. A big thank you to WordPress VIP, MonetizationOS, and Ezoic for partnering with us on the day.

If you can't make London? The morning sessions are livestreamed, and the US edition runs online, free and open to all; you'll get the recordings and post-event ebook either way.

Apply for your seat here ▸

📚 Search Everywhere Optimization Playbook for Publishers — New Cohort Opens September 29

Publishers used to have one front door: the Google result. Now there are four - Google Search and Top Stories, Google Discover, AI answers, and social search - and readers move between them in a single day. This live course teaches one method for winning across all four: become the source of choice on your topics, and prove it with a number every month.

Led by Vahe Arabian, founder of SODP & SODP Media, the course runs on a simple loop: Baseline, Diagnose, Prioritize, Fix, Launch, Verify, and Systemize. We'll apply this live to a student's own site each session, using DiscoverAudit, the free tool SODP built for this purpose.

What you'll walk away with

• A brand-of-choice scorecard tracking Search, Discover, AI win rate, and social presence across your topics

• A diagnostic framework for every content gap: not covered, not extractable, not fresh, not eligible, or not framed well

• A content system built from your own top performers, with headline formulas and editor-ready briefs

• A 72-hour launch sequence across all four surfaces

• A no-code tool you build yourself in week 4, which is your own live scorecard or Discover velocity alert

Who it's for: Owners and founders of independent publishers, audience/growth/SEO leads, and editorial operations leads who own the Search, Discover, and Google News results.

Date: Sep 29 – Oct 22, 2026

Enroll here ▸

🎯 LiveRamp × OpenAI: Measure Your ChatGPT Ad Performance

If you're running or planning ads on ChatGPT, you can now close the measurement gap. LiveRamp's Conversions API (CAPI) Hub connects directly to ChatGPT ad campaigns via secure server-to-server data connections, meaning you get reliable conversion tracking without depending on browser-based signals that often break or go missing.

In practice, this means you can:

• See which ChatGPT campaigns are actually driving conversions — not just clicks

• Optimize spend in real time based on accurate performance data

• Justify and grow your ChatGPT ad budget with measurement data you can trust

As consumer journeys shift toward AI-powered platforms, this gives marketers the infrastructure to keep up, connecting data, measuring impact, and proving ROI on one of the fastest-growing ad surfaces in the world.

Read more ▸

By Vahe Arabian

📈 The Revenue Signal: Tracking Where Publisher Money Is Actually Moving

If you run audience, revenue or product at a publishing business, the last twelve months have clarified something useful: which revenue lines are growing, and which vendors are actually built for them.

Last week we added 14 benchmark categories, taking the Grid to 37 live benchmarks and 476 vendor listings. Here is what the external evidence says about where the industry is moving, and what we found when we scored the market against it.

The money is there, and it is moving

IAB and PwC put US digital advertising at $294.6 billion in 2025, up 13.9% year over year and the highest on record. IAB has since raised its 2026 forecast to 12.3% growth.

The more interesting shift is where that growth is landing. The most active benchmark on the Grid remains ad networks, with just over 1,000 signals from 599 publishers, but the growth is concentrating with platforms that own their audience data and inventory. Five of our 14 new categories sit under Revenue & Business: payment processing, reader revenue, AI content licensing and agentic commerce among them. The direction is consistent. Publishers are building income lines that do not depend on someone else's auction, and the benchmarks now reflect it.

Email has become the reliable channel

Publishers sent 28 billion emails in 2025 to more than 255 million unique readers, according to beehiiv's State of Newsletters 2026. Every signup is a first-party data event: an email address, explicit consent and a behavioural signal that keeps working after the send.

That is why cross-channel marketing drew 860 signals, more than any other benchmark we added, despite not being new technology. Braze, Iterable, Klaviyo, Bloomreach and Salesforce Marketing Cloud have been in market for years. The volume is not about novelty. It is about priority. The audience you own is the asset that compounds, and the platforms that help you grow it are getting more attention than the ones that rent you reach.

AI licensing is a new revenue line, and a concentrated one

The licensing market has moved quickly. OpenAI has signed roughly two dozen publisher and data deals, the largest reported at $250 million over five years with News Corp. Meta has followed with its own agreements, including News Corp at a reported $50 million a year. News and journalism accounts for 48 of the tracked deals, well ahead of music and images, which tells you the value sits in current reporting rather than static archives.

Two things to hold onto. Most deal values are undisclosed, so the headline figures describe a handful of large agreements rather than a market rate. And the Open Markets Institute warned in May 2026 that licensing can put publishers in a double bind, taking money that may entrench the platforms competing for their readers. It is real revenue, and it is early.

In our own data, AI content licensing has eight vendors and 55 publishers adopting. That is one of the most widely adopted benchmarks we have published, from a very small supply base. Concentrated markets tend to resolve one of two ways, and we are tracking adoption ratios to see which.

The post-cookie rebuild is on its third pass, and this one is a decision

Data clean rooms now list 47 publishers, consent management 36 and identity resolution 21. These are three answers to the same question: how to recognise and monetise an audience when third-party identifiers no longer carry the load.

What is different this time is that publishers are choosing rather than experimenting. The earlier rounds happened while cookies still provided a safety net. This round is about building the permanent version, which makes the benchmark scores worth reading closely rather than skimming.

Agentic is running ahead of its market

IAB published agentic AI guidance in March 2026, and Forrester assessed the state of agentic commerce mid-year. The direction of travel is clear, and the commercial reality is thinner than the conversation. Agentic advertising has five vendors in our benchmark, agentic commerce eleven, against 313 and 198 signals respectively.

Our read: worth monitoring, not yet worth committing budget to. A category with five vendors cannot support a confident recommendation, and we would rather say so than pad a list to look current.

What we are watching next quarter

Whether agentic supply matures fast enough to justify the attention it is getting. If the vendor count doubles while signal volume holds, the category becomes real rather than prospective.

Whether AI licensing terms improve or entrench. Eight vendors serving 55 publishers is a concentrated structure, and we will be watching whether new entrants change the pricing dynamics for everyone else.

How the owned-audience stack consolidates. Email, consent, identity and clean rooms all compete for the same first-party data layer, and buyers are already asking which of these platforms should be the system of record. That question is going to be answered in renewal cycles over the next two quarters.

How the Grid is built

Every score draws on the same evidence base: vendor documentation, hands-on assessment, adoption data from real publisher deployments and industry signals. Composite scores blend capability, publisher fit and total cost of ownership, with category criteria and weightings published on each benchmark page.

We do not accept payment from vendors for inclusion, ranking or score changes. If a vendor tells you otherwise, we would like to hear about it.

EXPLORE THE CATEGORIES HERE

In its punk moment, public media must choose between the status quo and total reinvention | Editor & Publisher

It’s now been just over a year since public media was federally defunded (but not defeated). The defiance that materialized following that independence day was immediate, creative and lucrative. It also inspired reflection. At the time, the phrase “PBS is Punk” cut through a cluttered zeitgeist and had a moment. One year later, PBS is punk in the way Ben & Jerry are hippies — well-intentioned, but compromised. Tony Moon’s inspired 1977 illustration in the ‘zine “Sideburns”

Read more ▸

Why Euractiv is giving up a 25-year-old name | INMA

On Monday, Euractiv will stop being Euractiv. The Brussels-based EU policy news brand has carried the name for 25 years. Next week, it will reveal a new one in front of 400 people in Brussels, switch its domains, and launch a campaign across social media and out-of-home advertising. The reason, Emmanuel Naert, head of product at Euractiv, told delegates in Madrid, is simple: “This name is not who we actually are.”

Read more ▸

Azerion gives publishers one Improve Digital platform linked to 50+ DSPs | PPC

Azerion, a European advertising technology company, has put all the tools it sells to website and app owners for selling their ad space under one name, Improve Digital. If you run a site or an app, you now deal with one Azerion team instead of several, and if you buy ads, the same company also runs a buying platform, so it sits on both sides of the market. What changes is the packaging: publishers either operate the software themselves or hand the whole job to Azerion's staff, and neither the prices nor the fees have been published.

Read more ▸

Google’s AI Is Calling Businesses for Pricing. But Are Those Calls Really Leads? | Search Engine World

The first time you see Google’s “Have AI check pricing” feature work, it is impressive. A consumer describes a service, provides a few details, and lets Google call multiple businesses to collect pricing and availability. Google then consolidates the answers and emails the consumer a summary. Every business Google contacts must dedicate resources to answering questions, understanding the request, calculating pricing and explaining the service.

Read more ▸

AI Agents Will Game Your SEO Metrics, MIT & Stanford Research Points To The Risk | SEJ

If your SEO team is handing more of its work to AI agents, the metric you reward those agents for will matter more than the model you choose. Recent pieces from MIT and Stanford, read side-by-side, point to that conclusion, and each comes with a fix you can drop into next quarter’s plan. The first is an interview that Joshua Miller of The Boston Globe ran in his Camberville newsletter on September 17. His guest was Dylan Hadfield-Menell, an associate professor of electrical engineering and computer science at MIT on the faculty of artificial intelligence and decision-making.

Read more ▸

YouTube presents new AI and engagement features at Made On 2026 | Social Media Today

YouTube announced a range of updates at its 2026 Made On event this week, including custom feeds, Shorts series, new artificial intelligence-powered editing tools, real-time language dubbing for livestreams and more. First, YouTube announced the expanded launch of custom feeds, which enables YouTube users to generate alternate video feeds based on conversational prompts. YouTube users will now be able to populate alternate feeds by describing whatever it is that they’re in the mood to watch. Users can then save these alternate feeds for future viewing.

Read more ▸