ZwillMedia Agency

13/10/2026

A 2026 analysis found Perplexity cited content published within the last 30 days at an 82% rate. Six months out, that number drops hard, and it's a good reason to look at the quarterly calendar most of us still run.

The fix is less dramatic than people expect. Update the dateModified signal on your best pages, refresh the numbers, and put a year marker like 2026 in the headings. The same piece credits those year markers with roughly a 30% lift in citation rates. It's cheap work and it gets skipped constantly because nobody owns it.

If your top pages haven't been opened since spring, start there this week.

Full breakdown is linked below.

ChatGPT, Google AI Overviews, and Perplexity use completely different citation logic. Here's the 2026 data on how each platform sources information — and how to get cited.

13/10/2026

Small edit you can make today: put the year in your headings. Leapd's 2026 breakdown of how the major engines source information found that a visible 2026 in titles and headings improves citation rates by approximately 30 percent. Plenty of blogs we look at are undated, and if a reader can't tell when a post was written, a model probably can't either. Date every piece. Then refresh the ones older than 30 days before they fall off Perplexity's shortlist.

ChatGPT, Google AI Overviews, and Perplexity use completely different citation logic. Here's the 2026 data on how each platform sources information — and how to get cited.

13/10/2026

Perplexity holds onto citations longer than ChatGPT does. Scrunch and Stacker put the average stay for a Perplexity citation at about 5.8 weeks, compared with roughly 3.4 weeks for ChatGPT.

For us, that's good news for a page we already refreshed, and a warning for one we left alone. A citation earned in March may not be the one you're holding in October. So our quarterly review now starts with the top pages, then splits them into what needs a rewrite and what just needs new numbers and a current date.

For Perplexity specifically, freshness is the first thing I'd protect.

AI assistants cite content up to 25.7% fresher than organic results. See how content freshness affects AI citations and how often to refresh for AEO.

13/10/2026

Quick check before you lock in next quarter's plan. A 2026 citation report puts the overlap between ChatGPT and Perplexity at about 11 percent of domains. That means most of the sites one engine cites never show up in the other. That's the reason we'd split an AI visibility plan by engine instead of running one content calendar for everything. Pull your top five buyer questions, run them in both tools this week, and see which platform actually names you.

A 5W analysis of AI citation stats synthesizes 680 million citations across ChatGPT, Claude, and Perplexity to show how models source brand information.

13/10/2026

A line called AI search in your monthly report is measuring three different systems as if they were one number. ChatGPT, Perplexity and Google AI Overviews pull sources from different places, so a blended figure hides the exact part you need to act on.

We ask clients to keep one scorecard per engine. For each one, track where it cites your brand and where it cites a competitor instead. That second column is usually where the useful conversations start.

The harder part is the team meeting. Treating an old guide and a fresh forum thread as interchangeable is where most of the argument begins, and it's worth settling that before anyone plans the content calendar.

13/10/2026

Measure the search that happens after the answer, not the click.

If 93 percent of AI search sessions end without a visit, a referral report will undercount you almost every time. The buyer reads the summary, sees your brand named, and then searches for you directly. That branded search is where the business impact really shows up, and most dashboards never connect it to AI visibility. Put branded search volume next to your citation tracking and look for lift after you start appearing in answers. If you only watch one platform, you're blind to most of the picture anyway.

12/10/2026

Plenty of teams can point to one AI tool that's working. McKinsey reports that 88% of organizations use AI in at least one business function, but only 23% are scaling an agentic system.

That gap is where I'd look first. A chatbot that drafts captions is a tool. An agent that pulls campaign numbers, flags underperforming posts, and updates a client report on a set schedule is a process, and processes need an owner, a log, and someone checking the output.

If one person still has to remember to pull the numbers every Monday, you're probably in the first group, not the second.

12/10/2026

CrewAI surveyed 500 senior executives at companies with $100M+ in annual revenue to see where their AI agents actually stand. 65% say their organization already uses agents, and 81% say adoption is fully scaled or still expanding.

The number I keep rereading is 31%. That's the average share of workflows these companies say they've automated with agentic AI. The rollout is real, but most of the work underneath it is still getting done by hand.

For a lean internal marketing team, I'd turn that into a quick checklist. Which workflows do you own end to end, and which ones are just a pilot sitting in a slide deck?

65% of respondents said their enterprises are already using AI agents today; 81% reported adoption is fully scaled or actively expanding.

12/10/2026

Only 12% of CEOs say they get both revenue gain and cost reduction from AI, according to a new PwC release.

Most companies still score their agents on deflection rates and ticket counts. Those numbers look good in a deck, but they don't show whether a lead costs less to acquire than it did before the agent went live.

If you run marketing for clients, bring that question to the table. Ask what the agent changed in cost per lead. The companies moving past pilots tend to have rebuilt the workflow first instead of bolting agents onto the old process.

12/10/2026

Perplexity has committed a $42.5 million pool for publishers under its Comet Plus program. Subscribers pay $5 a month, and 80 percent of that goes out to publishers.

Before anyone on your team pitches AI citations as a revenue line, look at what hasn't been disclosed. Actual payouts, the weighting between visits, citations and agent actions, and the number of paying subscribers behind the pool are all still private.

We spend a lot of time helping clients tell which numbers are real and which are projections. This one is a commitment with a lot of blanks.

Comet Plus is Perplexity's $5 a month subscription that sends 80% of revenue to publishers from a $42.5 million pool, paid on human visits, search...