Kieran Flanagan’s Post

Here's what most people are getting wrong with AI They're not making the right bet You should build for the AI of the future, not for the AI of today It's a given these models will get exponentially better If you're using AI to accelerate your go-to-market, see where AI *should* make a meaningful difference, even if it doesn't today, and start building those use cases. The current models are a great way to test those use cases, even if they're not good enough to put that use case live. Once the new models come out, you can plug them in and instantly upgrade the performance of that use case so it can start creating impact for your business. AI Magic! It means you'll be first. And being first matters a lot with AI.

Google's search engine share in the USA fell to a 15-year low What happened? Google's search dominance in the USA took a hit in April, and its share fell to 77%; typically, it's 85% to 90%. Bing increased their share by 5% to 13% On Marketing Against the Grain, we predicted that people would start choosing an 'Answer' engine instead of a 'Search' engine. Bing was early to launch their AI copilot. Are we starting to see this trend play out? Or is it potentially a data anomaly or a large partner switching from Google to Bing? We'll break it down in detail on the podcast. We're early on all the trends impacting marketers. We first talked about the impact of the 'Answer' engine in November'22. Follow me here and the podcast to get them first. You either need to ride the trend or get rolled over by it!

The best marketers sharpen their skills by doing sidequests. Here's why. Too many marketers graduate to leadership positions and stop practicing the craft. All their time gets hoovered up by people management, strategy, and memos. We recently had Gary V on the podcast (episode out soon). He's worth hundreds of millions of dollars - he can hire people to do the craft; he doesn't need to do it anymore. But he's still doing it all: creating content, running ads, practicing, practicing, practicing—decades on from when he first started. Right before he came on the pod he was writing copy for instagram posts. I loved it. It reinforced my belief that the best marketing leaders are marketers first and people managers second. It's why I put out a free email course last week teaching marketers how to use AI. The course featured six lessons, six advanced AI prompts, six AI-generated videos, and a custom GPT—all done with AI. I learned a lot. My goal for it was tiny: to hit 1500 signups. It's at 1,100 so far. It's been so much fun promoting it, getting feedback from users, seeing how I could have made it better, and refreshing my HubSpot dashboard to see how close I am to my target. Side quests don't need to be giant goals. They can be small things to sharpen your skills. Always be practicing the craft. It's the fun part!

I had a blast working for Zapier. Finished now. Here are 3 learnings I added to my swipe file from my Zapier journey. First, a huge thanks to Wade Foster for having me. I've been a big fan of Wade's and Zapier since I first spoke with Wade on my podcast in 2017. I also loved my time working with Giancarlo 'GC' Lionetti, who was a great thought partner and a bunch of fun to work with. The entire Zap crew is great. I very highly recommend working at this company. You'll learn a lot and get to do real work. Onto the learnings: 1. Tactics vs. Strategy: Something I love about working with Wade is he wants to dive into the work, the details, the tactics. In my first presentation to Wade, I tried to dazzle him with a strategic memo, with the work being shipped, e.g., the tactics, as a footnote. My thinking was that Wade is the CEO; he won't want to see the granular details. But Wade wanted to see the details, and I loved this. It set the tone for all leaders to dig into the details, and I love the details, e.g., the actual work. Too often, especially in large companies, most of the work is regurgitating strategic memos, with little of that memo ever being actioned. The winning is always in the work. 2. Customer Obsession: Zapier obsesses over the customer. GC instrumented a GTM rule that everyone must listen to customer calls. You need to understand how customers experience problems. I remember when considering how to move a specific metric; I had a bunch of preconceived ideas on how to solve it by looking at the data and going through the experience. I then listened to 20+ customer calls where they mentioned challenges with the current experience, and some of the challenges they mentioned were things I'd never had thought of without listening to them. Wade constantly told teams to get out of the data, listen to customers and experience the problem like a customer. No matter the size of the company, staying close to the customer is a winning strategy. 3. Ship, Ship, Ship: Anyone who's paid attention to Zapier this past 16 months knows one thing to be true: they are shipping !! Zapier is a shipping machine. Part of this is asking if we can go faster. Are we being too precious about what's being shipped, at what point can we put something live and get feedback from customers. See point 2, customer feedback points us in the right direction. Part of this is asking if we are being ambitious enough. Are we setting an ambitious date? Are we pushing ourselves? Every tech market is now super competitive. Ask if our competitors would take this long to ship and put it live. It's a great forcing function :) It was very fun times at Zaps. Next Stop, I'm going back to the orange sprocket. All the best things deserve an encore, and there are some exciting projects I'll be doing back at HubSpot. It isn't surprising that a lot of it is AI :) Stay Tuned for updates on what the new AI Go-To-Market looks like—I'll be busy building it!

Here's how to create a marketing campaign in under 60 seconds with AI We'll include 7 influencing techniques from Dr. Robert Cialdini Here's another in-depth AI prompt. This time, it's focused on creative ideas for marketing campaigns. There's a technique in this prompt I've found useful. Creating a loop and asking the AI to iterate and edit its work. This is from a short AI course I put together for marketers. I released it on Monday for fun; over 800 people have signed up already. Pretty cool. Link here (replace DOT with .) https://bit[DOT]ly/timeless-marketing-ai The following prompt will create 5 iterations of a marketing campaign; each iteration should improve on the previous version. Quick note: This worked great on ChatGPT Pro some weeks back. Now, it's gotten lazy and will only do the first iteration. Just ask ChatGPT to complete the loop. It works great on Claude. I have to try it on Gemini Advanced (I'm doing it today). Full prompt here (replace DOT with .) https://bit[DOT]ly/marketing-campaign-ai

What does the AI Marketing Playbook look like? Four Pillars 1. Personalization at Scale: Marketers automatically default to email when I say this. But, think about lead magnets like ebooks and templates, all created for a single person. Landing pages customized to an individual's needs. Media designed for an individual vs. persona where the consumer gets to pick what's created for them. That's where we are heading. 2. Customer Journey Collapse: Today, the customer journey is a reflection of your org structure. Customers experience the handover points; it's messy. AI will allow single teams (I believe marketers) to own more of the customer journey, automate sales outreach, automate customer support. It will be a beautiful time :) 3. Multi-Modal Content: We default to text because it's cheaper and easier to create. AI will let marketers let their creativity run wild, easily bringing those ideas to life through audio and video. All media strategies will need to evolve to be multi-modal, and ideas will need to work across multiple formats. 4. Creator-Led Strategies: We're seeing a shift from brands to creators. Marketing channels that are growing favour individuals, not brands. Creators will become part of a B2B's marketing mix. The boom for B2B creators is only starting. ---- Lastly, marketers will fall back in love with the craft of marketing. AI will help automate much of the science, leaving the craft a place where differentiation is possible. I got to talk about this new playbook in detail with Brian Balfour and Fareed Mosavat on one of my favourite podcasts - Unsolicited Feedback. The episode is out now!

AI means businesses need to learn a whole new category of SEO. Optimizing for LLMs. Here's how you do it ... Ethan Smith appeared on the podcast and discussed his thoughts on optimizing for large language models, like ChatGPT. He talked about two things, which I found fascinating: 1. Listicle Optimization: For a product to appear on a best-of list, you'll need to ensure your product appears in as many listicles as possible on that topic. This is how LLMs will determine what should make the list. The cool thing is this was something we started at HubSpot years back. 2. Co-Word Citation vs. Link Building: Today, you build links. For LLM optimization, you'll build co-word citations where your product appears as the next word for the keyword you're optimizing for. We, of course, covered ways you can hack these things as well :) Ethan has an incredible thought exercise on how affiliate networks might work to implement co-word citation for brands. If you're concerned about AI's impact on SEO and how that effects your business, this is the episode for you. Out now on Marketing Against the Grain.

Marketers undervalue AI because they need inspiration on how to use it. Let's fix that with some advanced AI prompts for marketers. I created a fun short email course that turns 'Timeless' marketing lessons into advanced AI prompts. It includes marketing lessons like - brand stories Seth Godin would create, landing pages John Caples would write, headlines David Ogilvy would use, or entire marketing campaigns Dr. Robert Cialdini would produce. Each email teaches a single lesson from these greats, gives an example of that lesson in the real world, and an advanced AI prompt to bring it to life for your product/service. It also has a custom GPT to implement each lesson for any product/service you want. I'm not selling anything. I'm passionate about AI being a boom for marketing. My only ask is feedback from people who take the course on additional content you'd like on AI & marketing. Let's turn more marketers into CROs, CEOs, and founders. AI is a way to do it. You can signup here >> https://bit[DOT]ly/timeless-marketing-ai Replace the DOT with a . Oh, and there is a fun AI lesson once you join that I don't cover here. It's only for subscribers :)

ChatGPT launched Nov'2022. The next day, we did a podcast; I said this would cannibalize Google search. Why so confident? Google's entire mission is to organize the world's information and make it universally accessible and useful. Google wants to give users answers as quickly and accurately as possible. This is why Google launched "Quick Answers" in 2014, eventually becoming featured snippets in 2016. Google could now answer a user's query without that user having to click on the publisher's website. In HubSpot, we did a study on featured snippets and saw that they cannibalized search traffic. When they appeared, the keywords on the first page of Google got less traffic. Since then, the % of no-click searches (users who search for a query and don't click on any results) has grown to over 50%. Even before AI Chat appeared, a large percentage of users were already choosing to get answers without clicking on blue links. That behavior resulted in publishers getting less traffic to their rankings. When I first tried ChatGPT, I saw this was the endgame. AI Chat assistants like ChatGPT, Microsoft CoPilot, and Claude aren't a like-for-like replacement for Google's search engine. They're often confidently wrong. But they'll get more accurate, and you can validate the accuracy of most answers by asking them to cite their sources. They're better. Easier, faster. They're featured snippets on steroids. There is no turning back. User behaviour is a great predictor of what's to come. In this case, users have been telling us for some time that they'd prefer to be given an answer rather than trying to retrieve it on their own.

Scaling your marketing is incredibly hard. It's not for the reason most people think. The most common challenge for scaling marketing is saturating your main channels. Diversifying your channel mix when your top channels become saturated is very hard. But, here's another thing that happens. Your success is built on the work done by individual contributors. Sure, everyone is essential to your success: managers, directors, and everyone. But your ICs spend all their day doing work. A great IC is likely 3X as valuable as an average IC; a world-class IC is more like 5X/8X. We've taught people that to progress in their careers, they must move into management/strategy as fast as possible. It means you have great and world-class ICs telling you they'll leave unless they get to move into 'strategy'/'management' after only being in their role for 12 months. All managers should have deep expertise in their area, but that doesn't necessarily mean a world-class IC will make a great manager; they are different skillsets. That means you lose your best ICs and often gain ok-ish managers, and replacing those ICs is hard. Marketing leaders should prioritize creating an environment where the best ICs can stay as ICs for as long as they enjoy the role. Give them more autonomy, remove blockers, don't load them up with mixed priorities, give them big and interesting projects to work on, and let them go. At Zapier, I prioritized this with the marketing and growth teams. I spoke with every IC and tried to figure out how to make them the happiest cohort in my group for this exact reason. The fewer incredible ICs you lose to strategy as you attempt to scale your marketing, the more successful you'll be. We talked about this, why CMOs fail, who a founders first marketing hire should be and much more on the podcast with Jason M. Lemkin, episode out now! Go listen to Marketing Against The Grain