Kieran Flanagan

Today is my last day at HubSpot

Here are some of my favourite marketing tactics HubSpot pioneered to grow so fast

Some of these are commonplace today, but in the early 2010s, most B2B brands weren't doing this kind of marketing.

1/ Email Virality

I was amazed when I joined HubSpot that emails to HubSpot’s database generated 15%+ of new contacts.

Databases were for nurturing people to become qualified leads

HubSpot's content was so good people forwarded the landing page to colleagues, who downloaded it. We optimised the hell out of the share with colleague CTA's. Worked incredibly well.

2/ Own the internet for a day:

HubSpot thought big. Create an offer of such value; the internet would talk about it that day.

Like when webinars were cool (2010), breaking the Guinness world record for most attendees.

That obsession with value drove a lot of the early success.

3/ Market with Friends:

When I joined HubSpot, there was a team called 'comarketing.'

They were crushing it. They'd find partners with similar audiences, think of great campaigns and market together.

Today, comarketing is still a big part of how HubSpot generates demand.

4/ Templates:

Sure, today, everyone is doing templates.

But HubSpot was one of the first to create easy templates for marketers on things they needed.

Back in the day, when published, some of these template offers drove so much demand they crashed our systems.

5/ Thank You Pages:

Our thank you pages after someone downloaded an offer got SO much traffic.

We optimized the hell out of everything that got traffic. The quick wins.

Those thank you page optimizations drove a lot of MQLs.

Master the quick wins and place long-term bets.

6/ Be First:

HubSpot was one of the first B2B brands to lean big into social, especially Twitter.

They hired people like Dan Zarrella to create unique and engaging content about the platform.

Early adopters eat first

7/ The Story:

HubSpot's early pitch was so good

Make Marketing people loved

The story resonated so much with the target market; it was transformational. It's why I joined.

An early example of deck here ->
https://lnkd.in/eyRTH7hz

8/ Don't be Afraid:

HubSpot was never afraid to try new things.

Before joining HubSpot, I was a HubSpot TV fan.

It was a YouTube show before its time (2010).

Experimenting with new platforms, content, and stories is such a critical part of growth

---

People think of HubSpot’s content’s SEO & machine

But, HubSpot didn’t focus on SEO until much later. All of the above came first.

That’s what made HubSpot successful, they zagged when everyone else zigged.

What a journey!

Claude just released the best feature yet for writing

I've been doing it all year, and now they've made it available to everyone.

Here's a complete tutorial

Product marketing at Anthropic and OpenAI must be wild.

CPO: “We need to update our positioning. Our latest models are amazing. You can do motion graphics, build apps, solve real problems, help people with strategy.”

PMM: “Great. I’ll start repositioning AI as your strategic partner and always-on worker. Something that transforms a company from slow and cumbersome to fast and agile.”

CPO: “Great. One more thing. It also might kill all of humanity.”

PMM: “Sorry, wha—”

CPO: “Yeah, it’s incredible at writing. Can teach your kids amazing knowledge.”

PMM: “Yeah. Sorry. The killing thing?”

CPO: “It can automate tasks, help doctors cure diseases, help scientists solve problems never before possible.”

PMM: “Right. Yep. Amazing. Sorry, I just thought I heard ‘kill all of humanity.’”

CPO: “Yes. Exactly. It can do all those things. And it might kill everyone.”

PMM: “Everyone?”

CPO: “Potentially, yes.”

PMM: “Right.”

CPO: “Anyway. Great chat. Can we have the website updated tomorrow? And get it into the sales enablement decks.”

PMM: “What do I put under ‘Why now?’”

CPO: “Probably the productivity stuff.”

If you’re a marketing leader, here is what I’d build with AI

1. Today’s priority list - gather updates across all important channels and stack-rank your priorities by what matters; e.g., mine are blockers, KPI’s off track, follow-ups I owe, exec requests.

2. People brief - make 1:1s valuable by having AI create a brief for them; mine surfaces follow-ups from the last meeting, any blockers, challenges, and current KPI performance

3. What did I miss? - I have a skill that prioritizes messages I need to come back on across email, Slack, live meetings (via notes), and Google Docs (comments)

4. Exec feedback panel - voice of your execs to get feedback on pitches before the actual pitch

5. AI coach - a skill that looks at your AI usage and tells you how to improve. It can literally tell you what skills you should build based on usage

6. Voice of the customer - a skill that aggregates and builds a live artifact showing feedback across review sites and internal sales and customer calls. Add to a Claude/ChatGPT project to talk with your customer and get instant feedback

7. Voice of the competitor - a skill that aggregates and builds a live artifact showing updates to your customers' product pages and case study pages

8. Blocker watcher - a skill that searches across your Slack, email, and Google Drive to surface any blockers teams have and propose mitigation plans

9. Personal intelligence layer - a skill & artifact that continually enriches with your key learnings, failures, and what works. FYI, covering this on my Substack tomorrow

10. Voice of you - a synthetic version of you that your team can ask questions of before asking you

11. Storyteller - a skill that turns your updates and thoughts into incredible internal comms. The toughest part of your job is selling the value of marketing internally

12. Custom dashboards - a skill & live artifact to build a custom dashboard to keep track of how teams are progressing

13. Are we really different? - a skill to grade your product positioning against key competitors and grade how differentiated it is

14. Ad creative at scale - use AI to test way more ad creative across paid; this is where a lot of the leverage is

15. 1:1 sales enablement for enterprise - custom landing page, presentation, and video for large enterprise customers. Move from segment marketing to 1:1 marketing

16. Challenge me - skills that adopt different problem-solving frameworks and challenge your approach

17. Claude Record Skills - spend a day using Claude's new feature that allows you to record yourself doing work and auto-creating skills from it

The crazy thing about AI is we are just limited by our ideas. You can give the above list, as is, to ChatGPT or Claude and likely build out nearly all of these with just a tiny bit of back and forth.

The rules of marketing are changing faster than any other discipline right now …

Here are some of the new rules as I see them

- Brands dominate marketing channels → Creators dominate marketing channels
- Consistency and best practice win → Speed and learning win
- Don’t be controversial, stick to brand talking points → controversy sells, have something to say
- Best practice on known channels → New tactics on new channels
- 80% text; 20% video → 80% video; 20% text
- YouTube is for vloggers → YouTube is for everything
- Buy space on relevant media → Create and own relevant media

LinkedIn reach has declined a lot for people in July. I was curious what had changed. LinkedIn content analytics isn’t good. Using Claude & a custom dashboard, I got an answer why.

It’s estimated that AI now generates nearly 60% of posts on LinkedIn. It likely generates 99% of the comments (jk)!

It makes sense that LinkedIn is continually experimenting with its algorithm to figure out how to improve this.

LinkedIn’s content analytics sucks. And their export is missing many key metrics I’d want to track.

Quick edit. I had originally shown how I did this with Claude using a skill and their browser extension. Unfortunately, as people in the comments have pointed out, this isn't allowed, even if it's your data. So, I don't want to get anyone into trouble. My recommendation is to take screenshots of each post's analytics and have Claude parse the data to build this dashboard.

The dashboard gives me:

- Monthly trend charts for impressions, reach, followers, gained, and engagements.
- Click any month to see an overview of performance for that month, broken out at the post level.
- Top 10 leaderboards broken out by impressions, engagements, and virality.
- In-network vs. out-of-network reach so you can see which posts break out of your bubble. THIS is the most important metric for LinkedIn that no one tracks. The dashboard makes it easy.
- A 🔥 viral badge on posts where more than half the reach came from outside your network
- Full post content you can click to expand and re-read
- Claude chat so you can query the data

One insight Claude gave me is that LinkedIn has started to limit a post's reach to your own network (followers), then expand it gradually beyond that based on engagement.

I shared a templatized version of the dashboard on my Substack, so you can download the template and ask Claude to replicate it from the post images.

This week, I spent > 12 hours recording the audio version of our upcoming marketing book, Loop. It gave me a lot of time to reflect on the current state of marketing & AI.

First, this is such a full-circle moment for me. I often say my real marketing career started when I read the book on Inbound Marketing from Brian Halligan and Dharmesh Shah . Getting to write the next chapter of that with my great friend Kipp Bodnar has been super cool.

In the book, we talk about this being the most disruptive time for marketing, and yet, likely the most exciting. That’s the true paradox.

And what separates those two things is how “AI curious” you are.

I recently sat with a marketer who showed me a really cool AI workflow they’d built. Something that was real leverage for them. And guess what? I didn’t think they had any special talent that no one else possessed. They were just more curious than other people.

They showed me their backlog of how they built it. They asked AI, could you do this? Well, what about that, ok, that’s cool. Could we try tweaking this? Actually, I heard about this new AI tool. Could we connect to that… and so on and so on.

They didn’t stop at the second or third question; they kept going, asking, probing, trying to figure out how to make something genuinely great.

That’s it; that was their leverage. Something that is easy to replicate, but most people don’t.

The ‘AI curious’, the marketers who exhibit that trait, are the ones most excited by AI right now. Because how you win, which we attempt to document in the book, has changed so profoundly.

AND lastly, people, recording the audio version of a book is, I think, maybe, the most brutal, grind-it-out experience ever. Be warned if you ever decide to do it. It turns out, saying words is kinda hard lol.

Sometimes, it’s worth remembering that truly exceptional marketing was truly exceptional long before AI arrived. A lot of marketing today is instantly forgettable. What makes it timeless is


Creating value for customers long before they pay you a dollar, building a connection between you and the customer long before they ever speak with your sales rep, painting a picture of how that person’s life will be better once they use your product or service, and trailblazing a new tactic.

It’s …

- Rosser Reeves creating the strapline for M&Ms - “melts in your mouth, not in your hands” - in 1954, making it the longest-running slogan in advertising history.

- Paula Green writing "We Try Harder" for Avis in 1962. The company hadn't made a profit in 13 years. Instead of hiding the fact they were second to Hertz, she leaned into it: "When you're only No. 2, you try harder." Within 12 months, Avis went from a $3.2M loss to a $1.2M profit.

- Dan Wieden coining "Just Do It" for Nike in 1988. Nike was an $877M running shoe company. The slogan turned it into a lifestyle brand. Sales hit $9.2B within a decade. Market share jumped from 18% to 43%.

- Brian Halligan and Dharmesh Shah coining the term "inbound marketing" in 2006 and building an entire movement around it. The brand was the ‘feeling’ the methodology gave you.

- Canva scaling an SEO engine by turning their product into their content engine via thousands of templates. It now drives 73M+ organic visits a month and helped build a $40B company.

- Red bull deciding to sell its energy drink by becoming a media company, through a network of YouTube channels, sporting events, and documentaries.

- A Spotify intern named Jewel Ham designing shareable Instagram Story cards for Wrapped in 2019. It turned 60 million users into unpaid advertisers in a single week, and the app jumped from 15th to 2nd in the App Store overnight.

- Duolingo handing TikTok to a tiny in-house team with full creative freedom and zero corporate polish. They turned their owl mascot into an unhinged, meme-fuelled character that never once sells the product. Result: 50K to 16M+ followers, monthly active users tripled from 40M to 116M, and they even "killed off" the owl in a stunt that spiked brand mentions by 25,000%.

Great marketing is still great marketing. AI hasn’t changed that.