For years, marketing followed a simple plan where a team set a strategy, built a campaign, launched it, and reported on the results once the quarter ended. In their UNBOUND 2026 session, HubSpot CMO Kipp Bodnar and SVP of Marketing Kieran Flanagan showed how AI broke the old marketing playbook at both ends, because buyers now get answers from AI tools before they ever visit a website, and marketing teams can produce blog posts, emails, and ads in a fraction of the time it used to take. Watch the full session below to hear their take firsthand.
With content so easy to make, the real challenge is choosing what to make, giving AI accurate customer data to work with, and learning from every launch fast enough to improve the next one. Getting all of those pieces to run as one connected system is called orchestration, and it is the next big problem marketing teams need to solve.
Solving it starts with the systems underneath your marketing, and that is the work Campaign Creators does every day. As a HubSpot Elite Solutions Partner, we help B2B companies set up their CRM, move data from older platforms, connect their tools, and get ready for AI, along with improving conversion rates and building the web pages buyers actually see.
The old playbook moved in a straight line from strategy to campaign to launch to report, and it depended on two things that are no longer true: that buyers would come to your website and that results could wait until the campaign was over.
Buyer behavior changed first, since people who see an AI summary at the top of Google click a regular search result in only 8% of visits, compared with 15% when no summary appears. That cuts the first step of the old playbook nearly in half, and even HubSpot, the company whose founders coined the term inbound marketing, felt it when marketing-generated leads that once drove 80% of its sales team's revenue stalled in 2024.
The reporting step is breaking too, and Kat Warboys, HubSpot's Senior Marketing Director for Asia-Pacific, has said that waiting six to 12 months to report on a campaign no longer works. By the time that report is ready, buyer habits, ad costs, and search results have already shifted, so the team has spent months acting on old information.
The funnel is turning into an hourglass because fewer people reach your website while the ones who do arrive are closer to buying. HubSpot CMO Kipp Bodnar and SVP of Marketing Kieran Flanagan used this shape at UNBOUND 2026, HubSpot's annual conference, to explain how a company can get fewer website visits and still grow.
Buyers now discover brands through AI answers, creators, online communities, and social apps built to keep people inside them, so fewer of those moments end with a website visit. 58.5% of Google searches in the US ended without a click on any result.
The buyers who do click through have often researched with AI tools, read reviews and peer discussions, and narrowed down their options already. Semrush found that the average visitor from AI search is worth 4.4 times as much as a traditional organic search visitor based on conversion rates, and AI also makes it easier to tailor the page, email, or offer each of those buyers sees.
Brands show up in AI answers through answer engine optimization, or AEO, which is the work of making your content and reputation easy for tools like ChatGPT, Gemini, and Perplexity to find, trust, and cite. AEO builds on the same basics as SEO, such as a sound website and genuinely useful content, and it puts extra weight on what the rest of the internet says about you.
The data backs this up, since an Ahrefs study of 75,000 brands found that mentions of a brand across the web had the strongest connection to appearing in Google's AI Overviews. Brands in the top quarter for web mentions earned over 10 times as many AI Overview mentions as brands in the next quarter down.
Warboys has described the same shift at HubSpot, where authority used to mean earning backlinks and now means getting mentioned on the sites AI trusts, such as Reddit, LinkedIn, review sites, and online forums. A few habits help a brand earn that kind of visibility.
Visibility in AI answers then feeds the rest of the loop, because the buyers it brings in arrive already familiar with your brand and what it stands for.
Worth a read: How to Benchmark Your AI Visibility Against Competitors Using HubSpot AEO
When content becomes cheap to create, the most valuable skill shifts to deciding what is worth creating. HubSpot CMO Kipp Bodnar and SVP of Marketing Kieran Flanagan pointed out that producing the work was the hardest part of marketing for about a hundred years, and now that every team has the same AI tools, a lot of content has started to sound alike.
AI-written articles briefly outnumbered human-written ones in November 2024 and have stayed about even since, according to Graphite research. Yet the same research found that people wrote 86% of the articles ranking in Google Search and 82% of the articles cited by ChatGPT and Perplexity, which shows that publishing more does not automatically mean getting seen.
Their UNBOUND session summed it up in one line: "When the cost of production is free, judgment is the only thing that has a price," and that judgment comes down to a handful of decisions a person still needs to make.
Teams that make these decisions on purpose publish less filler and get more value out of every piece they put out.
Taste, meaning your brand's point of view and its sense of what good work looks like, becomes a business asset once you write it down, keep it current, and feed it into every AI tool your team uses. Left alone, AI tends to produce something close to the average of everything it learned from, which is exactly where your competitors already sit, so a clear point of view is what pulls your content away from the crowd.
Bodnar and Flanagan call this written guide a taste profile, which describes who your customer is, how they talk, what earns their trust, and what makes them stop scrolling, along with your brand's stories and opinions. HubSpot offers a free Taste Profile builder if you want a template to start from.
What separates a taste profile from a style guide nobody opens is upkeep. HubSpot keeps its brand guidelines current by feeding them customer surveys, call transcripts, support tickets, and input from leadership, and Warboys has warned that AI scales whatever you give it, so a voice you never define ends up being defined by the algorithm.
A useful taste profile usually covers four things that keep AI output on track.
Treat the taste profile like a product your team owns, and update it whenever a test shows buyers responding to a new message so every AI draft that follows starts from a better brief.
Running marketing in two-week sprints gives a team 26 chances a year to learn what works, compared with four from quarterly campaigns. Kipp Bodnar and Kieran Flanagan, who co-host the Marketing Against the Grain podcast and co-wrote the book Loop, built their UNBOUND case for shorter learning cycles on that simple math and on a story from early aviation.
As Bodnar told it, Samuel Langley spent years and a large government grant on a flying machine that crashed into the Potomac River on its second test flight in December 1903. Nine days later, the Wright brothers flew at Kitty Hawk on a budget of about $1,000 after building cheap gliders and learning from every failed attempt.
A marketing sprint works the same way, and each one needs a few basic parts.
That written note is what lets each sprint build on the one before it. As Warboys has pointed out, test results that live only in people's heads leave the company when those people do, so logging every result in your CRM or a shared test library turns a year of sprints into lessons the next campaign can use.
AI speeds up the marketing loop by taking over the busywork between launching something and deciding what to do next, including the tasks Bodnar said used to make a two-week pace so hard to keep up, such as pulling the data, writing the recap, and suggesting the next test.
That busywork eats a lot of time, since knowledge workers spend 58% of their day on "work about work" like status updates and hunting for information, according to Asana's 2023 Anatomy of Work Index. The table below shows where AI gives that time back during a sprint.
|
Task |
Without AI |
With AI |
|
Pulling performance data |
Asking an analyst and waiting days or weeks |
Marketers look up results themselves within hours |
|
Writing the recap |
Spending hours building slides |
AI drafts a summary from the campaign data |
|
Making versions for each channel |
Writing a new brief for every channel |
AI turns one proven piece into an email, video, landing page, and social post |
|
Personalizing outreach |
Adding a first name to a template |
AI writes each message using CRM records and call notes |
|
Choosing the next test |
Debating it at the monthly meeting |
AI suggests the next tests, ranked by likely impact |
The payoff can be large, as HubSpot saw an 82% jump in email conversion rates after using AI to personalize the follow-up emails it sends to new leads, and early users of HubSpot's Marketing Studio create 81% more campaigns on average.
Speed only helps if a person still decides what the numbers mean, so let AI draft the recap and rank the tests while your team makes the final call on what to try next.
First-party data, the information you collect directly from your own customers and prospects, carries more weight because AI can only be as good as the information it works with. HubSpot CEO Yamini Rangan shared an analysis at UNBOUND 2026 of AI results across more than 300,000 HubSpot customers, which showed that the same AI model running on bad context, such as messy or incomplete data, cut marketing-qualified leads by 28%, deals won by 27%, and meetings booked by 49% compared with using no AI at all.
Good data flips that result, with HubSpot Professional and Enterprise customers who use AI on complete, accurate business data generating 3.6 times as many marketing-qualified leads and winning 3.2 times as many deals as customers who don't use AI. That gap helps explain why roughly 90% of companies now use AI while only 6% say the results have been transformative, according to Rangan.
For marketing, the most useful first-party data comes from a handful of sources.
The same rule applies to paid ads, and as Flanagan pointed out at UNBOUND, an ad platform told to find form fills will find more form fills, while one given revenue data from your CRM learns to find buyers who actually sign. Tools that record calls, emails, and meetings automatically, such as HubSpot's self-updating Smart CRM, also help by keeping records current without relying on people to enter everything by hand.
A marketing tech stack, meaning the set of tools your team uses, learns when the results of every campaign flow back into the customer records, audience lists, and lead scores that shape the next one, while a stack that only sends emails and stores reports keeps doing the same things over and over.
Plenty of companies are not there yet, since Gartner's 2025 Marketing Technology Survey found that marketing teams use only 49% of what their marketing tools can do, even though those tools make up nearly 22% of total marketing spend. A stack that learns tends to show four habits.
None of this happens unless your systems are connected and your team uses them every day, because a stack with a separate webinar tool, a spreadsheet of event leads, and a CRM half the sales team avoids will always learn from incomplete information. Moving to one connected platform helps, but the bigger fix is usually process, which means someone has to own how data gets entered, how fields are defined, and how results get recorded after each sprint.
When traffic drops, marketing teams should measure the value of each visit and how quickly they learn, since traffic alone no longer tells the full story. These metrics give a clearer picture of how your marketing loop is actually performing:
|
Metric |
What it tells you |
|
Pipeline per visit |
How much sales opportunity each website visit creates |
|
Conversion rate by source |
Which channels, including AI search, bring buyers who take action |
|
AI visibility |
How often AI tools mention or cite your brand for buyer questions |
|
Learning speed |
How many tests the team finishes and acts on each month |
|
Revenue by channel |
Which channels lead to closed deals |
Traffic still belongs on the dashboard as a supporting number, and pairing it with the metrics above shows if the loop is producing real growth.
Marketing orchestration means coordinating your data, content, channels, people, and AI tools so every action builds on the last result. Traditional campaign management puts assets on a calendar, while orchestration runs a system where a signal, like a new question buyers are asking AI tools or a sudden jump in demo requests, sets off the next step.
HubSpot's platform is now built around this idea, and at UNBOUND 2026 the company described it in three layers, with Context holding what the business knows about its customers, Action covering the work AI agents and teams carry out, and Coordination deciding how people and AI agents share that work. Here is what one orchestrated loop could look like in practice.
Research shows the hardest part is redesigning how the work gets done, since McKinsey's 2025 State of AI survey found that the roughly 6% of companies that credit AI with at least 5% of their earnings stand out for redesigning their workflows. Even so, no more than 10% of respondents in that survey have scaled AI agents in any single business function, so plenty of teams are still early in this shift.
Orchestration also needs clear decision rights, so decide which actions AI can take on its own, which need a person's approval, and who makes the final call on budget, brand, and messaging.
The new marketing operating model works like a loop that keeps running, cycling through six steps called Express, Tailor, Amplify, Learn, Adapt, and Repeat, where the old model moved once in a straight line from strategy to report. These six steps build on the four stages of HubSpot's Loop framework, with its final Evolve stage split into learning from results and adapting the next cycle.
A loop like this needs a strong foundation, which is where Campaign Creators' Inside-Out framework comes in, since each of its five layers gives the loop something it can't run without.
|
Inside-Out layer |
What it does for the loop |
Signs it needs work |
|
Data Foundation |
Gives the system accurate information about customers and results |
Teams argue over whose numbers are right |
|
Architecture |
Connects that information to your processes, workflows, and handoffs |
Event leads, product data, or ad results sit in tools that never sync with the CRM |
|
Platform Adoption |
Gets your teams using the system every day |
Sales reps track deals in spreadsheets, and marketers export data to build reports |
|
Connected Experience |
Links every touchpoint in the customer journey |
Buyers repeat details to sales that marketing already collected |
|
AI + Optimization |
Creates the feedback loop that improves each cycle |
Test results stay in slide decks and never change the next campaign |
AI makes the marketing operating model more important, because it scales whatever your systems feed it and exposes a weak layer faster. The best place to start is your weakest layer, since a loop only runs as fast as its slowest part.
A continuous marketing loop depends on clean data, connected systems, and a team that uses them every day, and Campaign Creators helps B2B companies build that foundation in HubSpot, from CRM setup and data migrations to revenue operations and AI readiness.
The easiest way to begin is with a quick check, using the HubSpot AI Readiness Scorecard to see how ready your HubSpot portal is for AI, the Architecture Blueprint to map where your current tools break down, and the AEO Auditor to see how visible your brand is in AI answers. Once you know where the gaps are, our team can help you close them.