digitalNow 2026 wrapped on Wednesday, October 28, at the Hilton Arlington Rosslyn at The Key, just outside Washington, D.C. Eight keynotes in three days, plus more than 100 attendees who built with AI in the hands-on digitalLab. At the end of the conference, Blue Cypress CEO Johanna Snider pulled it all together the way David Letterman would have: a Top 10 list of the things attendees would be saying when they got back to the office.
Here is that list, with the stories behind each one.
HubSpot CMO Kipp Bodnar shared that HubSpot's organic traffic dropped by 80% while demand rose 20% over the same period. People didn't go away. They changed how they buy. They do their research in ChatGPT or Claude, and by the time they reach your website, they are coming to confirm, not to learn.
And the 404 page? In a single month, AI tools made up roughly 300 HubSpot pages that didn't exist and sent people to them, so an error page suddenly matters. Raja Rajamannar saw the same thing at Mastercard: if AI doesn't know the answer about you, it will make one up. Kipp's advice was to publish the things you'd rather not publish, so AI has the real answer.
Eric O'Neill spent his career catching spies, including Robert Hanssen. He expects that by the end of this year, about 90% of what we see online will be synthetic, and one fraudulent email can cost an organization $25 million. When anything can be faked, verify requests through a channel you already have, not the one the request came in on. Or agree on a safe word.
Raja Rajamannar, former Chief Marketing and Communications Officer of Mastercard, told attendees that 85% of people aren't hardwired for loyalty, even though companies spend about $1 trillion a year on loyalty programs. Points for dollars don't create loyalty. What you are really after is stickiness, and that is a different discipline.
Amith Nagarajan made a similar point about associations. Most members have an episodic relationship with their association: they join, attend a conference, take a course, and then go quiet until renewal. Sticky is what happens in between.
Career diplomat Danny Stoian shared how the Center for Emerging Global Leaders handled 300 fellowship applications without looking at college, GPA, or resumes, so every applicant got an equal shot. The real constraint was time, so they changed how the work got done. Every eligible applicant got a 15-minute, audio-only AI interview built around the five traits the program cared about most. The AI did the first pass, people checked its work, and they found great candidates in places they never would have looked.
Danny's point: "to AI or not to AI" is the wrong question. Start with the ambition, find the actual constraint, and keep humans accountable and in the loop.
Dave Dellin, CEO of the American Quarter Horse Association, follows a simple rule: outcomes first, AI second. At the American Paint Horse Association, which he previously led, about 30% of phone questions were already answered on the website or in the rulebook. APHA set a goal to cut routine calls by 30%, launched a chatbot named Becca, and reached a 28% reduction in the first month. They didn't start with the chatbot. They started with the outcome, and the chatbot was the tool that got them there. Read more in our digitalNow 2026 Day 2 recap.
Kipp and Raja both landed here. Everyone has the same models, tools, and prompts, so much of what comes out looks the same. Raja found more than 170 ad campaigns running at the same time built around the same iceberg image and nearly the same concept.
Kipp's point is that creating content is now basically free. Anyone can produce a hundred versions of anything in seconds, so the hard part is choosing what's good. His fix is a Taste Profile: write down who you are and how your members actually talk, so the AI sounds like you. Raja's version is that in a sea of sameness, you stand out on the human things: trust, empathy, and inspiration.
Safi Bahcall, author of Loonshots, told the story of Circuit City and Best Buy. Circuit City's board deeply believed customers wouldn't buy service plans without a salesperson but never tested it. Best Buy tested it. Best Buy is a $40 billion company today, and Circuit City is gone.
Kipp told a similar story about the race to fly. Samuel Langley had $1.5 million and crashed into the Potomac twice. The Wright brothers had about $1,000, built cheap gliders, and when they broke them, they fixed them and flew again. The challenge for every association: try one new thing a day, even something small, and when an experiment fails, learn from it and celebrate it.
Amith described run and build as two different jobs. The people keeping the lights on should stay focused on keeping the lights on. Ask that same team to also build what's next, and the day-to-day work will win every time.
Safi called it loving your artists and your soldiers equally. You need people excited to try things that might fail, and people whose whole job is to make things work the same way every time. When Steve Jobs returned to Apple, he had Jony Ive and he had Tim Cook. Without both, there would be no Apple today.
Sidecar's Mallory Mejias made this case in the final keynote, Auditioning for Relevance: Winning a Recurring Role in Your Members' Lives. For most of their history, associations were the only credible source of information in their field, because producing credible information was expensive. That isn't true anymore. Members can now choose from free AI assistants, vendors with content budgets, one-person newsletters, and their own peer networks. Your work didn't get worse. The room got fuller.
Drawing on hundreds of acting auditions, where competence gets you in the room but never gets you the part, Mallory broke down three things associations can use: understanding the situation members are actually in, an honest audit of what is still uniquely yours, and the trade-offs that separate a real decision from a template.
Safi asked the same question from a different angle: what can you do that no one else can? AI can write the report. It can't convene your industry.
Amith Nagarajan opened the conference with a slide of the five most powerful people in AI and a question: who would you trust to safeguard your association's data for the long run? His answer was to own your data, rent the intelligence, and control where AI goes in your organization. He also gave attendees four questions to ask any partner:
What tied the whole conference together, Johanna said, was trust. Eric O'Neill called trust an uncommon commodity, and it is one associations have in spades. Mallory talked about credibility that took decades to earn. That trust is the most valuable thing associations have, so build securely and build on a foundation you own. Read more in our digitalNow 2026 Day 1 recap.
digitalNow has always been about leaving with a plan, not just a page of notes. This year, more than 100 attendees went a step further and built with AI themselves in the hands-on digitalLab, run by the Blue Cypress Tech Fellows. Learning by doing is the fastest way to move from talking about AI to using it, and it is a model associations can borrow for their own staff and members.
Thank you to every attendee, keynote and breakout speaker, sponsor, and team member who made digitalNow 2026 happen. digitalNow 2027 heads to Chicago, November 14 to 17, 2027, and registration is already open.
digitalNow 2026 took place October 25 to 28, 2026, at the Hilton Arlington Rosslyn at The Key in the Washington, D.C. area. It is Sidecar's annual conference at the intersection of AI and associations.
The eight keynote speakers were Amith Nagarajan (Blue Cypress), Kipp Bodnar (HubSpot), Raja Rajamannar (Mastercard), Eric O'Neill (The Georgetown Group), Dave Dellin (American Quarter Horse Association), Danny Stoian (Center for Emerging Global Leaders), Safi Bahcall (author of Loonshots), and Mallory Mejias (Sidecar). Blue Cypress CEO Johanna Snider closed the conference.
According to HubSpot CMO Kipp Bodnar, buyers now research in AI tools like ChatGPT and Claude before visiting a website, so they arrive to confirm rather than to learn. AI also invents answers when it can't find real ones, so associations should publish clear, accurate information about themselves.
Eric O'Neill recommends verifying any sensitive request, like a wire transfer, through a channel you already have rather than the one the request came in on, or agreeing on a safe word in advance.
A Taste Profile, as Kipp Bodnar described it, is a written description of who your organization is and how your members actually talk. It helps AI tools produce content that sounds like you instead of like everyone else.
Sovereign AI is the approach Amith Nagarajan laid out in the opening keynote: own your data, rent the intelligence by treating AI models as interchangeable, and control where AI goes in your organization, rather than moving those assets into a single vendor's platform.
digitalNow 2027 will take place November 14 to 17, 2027, in Chicago.