How to give a killer product talk
Blog Post

How to give a killer product talk

Last edited: Aug 22, 2022 Published: May 5, 2016
Mixpanel Team

We’re not huge on listicles over here, but we’ve been told they’re engaging, and this article is all about engagement. Not in-app, but on-stage.

There may come a day when you’re invited to a sensibly lit room with sensibly adult beverages (such as Mixpanel’s Office Hours) to explain to a crowd of developers, product managers, data scientists, and tech enthusiasts how you built your product. And having a 20-minute mark to make your point, you’ll want to find the best way to communicate your biggest victories.

We’re going to help with that.

For the past few years, we’ve had the brightest minds in tech coming through on a (roughly) monthly basis to talk about how they used data to build awesome products. We’ve seen speakers from startups, mid-market companies, enterprise businesses, and incubators. They’ve called on all sorts of multimedia, sketches, and customer feedback, but the talks usually boil down to one thing: how we think about problem-solving.

It may seem simple and formulaic, but there’s actually an art to a fantastic product presentation. Sometimes even the slickest numbers are lost in translation. So, we looked through our archive and came up with some of the best strategies for talking up your product and hooking your audience.

Here’s what we learned.

Bear patrols and Boromir

“Another thing I want to talk about is something internally we call bear patrols.”

As it happens, Cotap is a messaging app, not a wildlife retrieval service. Everyone leans forward in their seats. What the hell is a bear patrol?

There was probably a time in your life when someone pulled you aside (or shared one of these listicle things) and tried to programmatically upgrade your public speaking routine.

“Insert a joke, and then pause for laughter.” “Use a visual aid.” “Find a good metaphor.”

Said in such a way, these tactics feels impersonally rote or glib. But there’s something to them. If you’ve included a boring point in your talk that needs spicing up, that probably means it was indispensably relevant in the first place. So, you want to make these instances of your talk memorable some way or another.

One way to introduce levity or personality could be borrowing some of your company culture or an in-joke.

When Cotap gave their talk here, they introduced us to the bear patrol. Cotap isn’t just interested in launching a feature and seeing whether the analytics tell them something good or bad. Like any proficient data-driver, they want to ask their data why something happened, so they can confirm that it’s related to whatever change they implemented.

This logic probably could’ve been bracketed under “correlation doesn’t equal causation,” but then-CEO Jim Patterson instead gave us a humorous glimpse into Cotap’s culture. When their team sets out to prove a trend is connected (or not) with an action of theirs, it’s called bear patrol.

The moniker comes from a Simpsons episode in which Springfield establishes a Bear Patrol. Homer “explains” to Lisa that, because there are no bears in Springfield, it means the Bear Patrol is working.

In Jim’s example of Bear Patrol, Cotap saw a dip in their numbers, segmented those numbers based on region, attributed the dip to Asia, and correlated the dip with the occurrence of Chinese New Year. “We call it seeing if there actually is a bear or not,” Jim said.

When Social Capital partner Ashley Carroll spoke at Office Hours on behalf of DocuSign, she used a similar beat. Partway through her talk, as she introduced the topic of taxonomies, she hit her audience with a picture:


The Lord of the Rings nerds are plentiful at our talks, and this definitely got them going.

At DocuSign, the engineers passed this meme back and forth when discussing implementation. The audience identified with Boromir. Now, they associate the lesson of a wise implementation with a strategic Mordor trip.

A meme is easy points, but points nonetheless.

Don’t shy away from little moments of levity like this. You want your talk to be memorable (and “human,” as we’re often reminded), and a meme goes over better than a mnemonic. We don’t call this cheap. We call it being effective.

Go in with a mission

“…And we’re hiring.”

Maybe it’s unique to San Francisco, or the startup world, but when we invite people to speak at Mixpanel, they generally make a recruiting plug somewhere in the second half of the talk. The plug sneaks in quick and sheepishly, but it often includes the office location and promise of a welcoming culture. A tad aggressive, sure, but I have a theory about the upside of this recruiting practice.

It forces the speaker to give enough context throughout the rest of the talk, if only to ensure the pitch works.

The startups that make this plug want prospective hires to understand their product and their problems. They trumpet not only their resounding victories, but also what they make and what sorts of problems they run into.

What is it actually like to be a data scientist at Hinge? A developer at Uber? A product manager at Postmates?

These speakers want to lure their audience into the laboratory, in the hopes of convincing the brightest to join them there. The strongest product presentations do double-time as informational interviews.

At a talk with a bigger name, the rep generally fails to do this. The giants assume you know their problems or else don’t care. Hiring might not be top of mind.

But they really should care.

Every time you speak in public, you’re fighting an uphill battle for attention, whether you’re a tiny startup or Elon Musk (well, he might have an advantage or two). A speaker is trying to win the audience, whether they choose to acknowledge it or not. Hungry to hire, smaller startups are correctly humble and know the fight is uphill. They provide ample context, as every speaker should.

The next time you give a product talk, imagine your ideal recruit is somewhere in the audience. Throughout your speech, angle to make her eager to sign your offer letter. The battle is not won on the familiarity of your logo alone. So, go in with the mission to evangelize yourself and your product.

You’ll be more engaging for it.

Show your best practices at the far limits

Find a widely applicable problem made uniquely harder by the nature of your product. That has the makings of a great story.

Ishan Verma came to Office Hours when he was a business analyst in AutoDesk’s Consumer Group. He presented recognizable problems, but at a scale unfamiliar to most in the audience:

“We had to create an event taxonomy. Maybe if you work for a small startup in the Bay, everyone’s in the same office, it’s easy. Our web guys are in Shangai. Our Android guys are in Israel. Our iOS guys are in Toronto. And we’re here. We’ve got three different platforms, four different time zones.”

Working for a firm as wide-reaching as AutoDesk, there were a lot of extremes. Another speaker might have generalized their talk for the minnows in the audience. After all, not many companies can flex three platforms in four time zones. Ishan, however, was smart to include this. It made his talk authoritative and memorable.

But Ishan was even smarter than that. He chose a business problem that people could relate to, regardless of AutoDesk’s size. Small startups in the Bay still create event taxonomies.

Even at the far limits of a product like AutoDesk’s, Ishan had best practices anyone could emulate. But because he showed them working at the extremes, their validity as best practices was confirmed.

You’ve got a product-market fit, an industry, a specialty. When you get on-stage, you may feel tempted to tell a cookie-cutter story. Don’t. Think specifically about your product and its challenges.

Devin Markell, who works in data at the popular dating app Hinge, adopted a similar tactic.

Re-engaging customers is always risky business. Without the right etiquette, it can come off as nagging. But in Devin’s case, re-engagement can leave a particularly bad taste in a user’s mouth:

“We may send a re-engagment email, basically saying ‘Come back to the app, we found someone great for you.’ They write back and say, ‘First, you find me my girlfriend, and now you want me to break up with her?’ which is kind of a spot-on idea of what we’re trying to avoid.”

Marketers and product managers the world over should be thinking about re-engagement. But Devin’s audience will remember this because of Hinge’s funny blunder, even if the stakes aren’t as high for them.

What problem has your product tackled against long odds? Sharing that story will not only make your team and your product interesting. It’ll also give your audience something to work with. We can all shore up our event taxonomies and our re-engagement efforts and probably have an easier time of it than Ishan or Devin did.

Mistakes were made

Wellll, we looked at the performance of those tours and I’m sure many of you can guess. After the second slide, everyone had dropped out of the funnel.”

Product owner Tara Pugh at Cozi was walking us through the failures of their funnel. Her team had designed some fairly congested UI. User drop-off may not be what you’re aching to brag about, but Tara knew there was greater insight to be gained by documenting the mistakes.

Giving a product talk is essentially storytelling, and there is no story without conflict.

In other words, you’ll bore your audience if whatever insight you present is a painless eureka moment your team alighted on one fine day.

Besides, that’s not how good products get made. In reality, product managers are just as often “problem managers”, and there’s no shame in being honest about that. Honesty fosters authority. When you own up to mistakes, people get that you aren’t bullshitting them. It humanizes. It’s relatable. Plus, it sets up a greater contrast for your eventual success.

When acting as an ambassador for your company or startup, it’s tempting to advertise your personnel as brilliant developers and PMs who never make mistakes. But this probably isn’t the case. Reserve that narrative for your careers page, not a conference.

Because, even if the solution you reached seems deceptively simple in hindsight, getting to that state is a lot of work. Your peers want to see that work. And as every clickbait-y listicle has proven, people want to learn about the One Thing That Will Shock You. Use that to your advantage!

Tara revealed the various A/B tests she and her team did, and what appeared as a deceptively simple UI was actually thoroughly-tested beauty. It was even more impressive to look at for having seen its transformation.

Admit your mistakes and your audience will venerate you for it.

Be directionally correct

“This is fake data, but it was pretty bad.”

That’s two engineers from Uber, Zach Heller and Kevin Roth, throwing up a funnel on the screen. After the first step in the funnel, drop-off was huge.

In the way of peer review, fake data is not perfect. But for illustrative purposes, it works, and the Uber guys got a laugh out of the crowd.

We pride ourselves on being data-driven here and hold our guests to such a standard. But often, data is proprietary and super-secret. Sharing hard numbers is rarely an option. The temptation might be to circumvent any talk that will require data, but vagueness doesn’t work. So instead, show how the data looked directionally.

At a decacorn like Uber, there’s a lot of red tape. But their engineers were scrappy. They knew how to navigate the stipulations imposed upon them. They turned what could’ve been a boring confidentiality disclaimer into a value-add.

Your audience isn’t asking for your secret formula. Just specificity. So if you can’t give them those cold, hard numbers, at least be directionally correct. Convey the stakes of your situation.

Negative space

Back when we were in smaller digs, there was no moody purple lighting, iPad surveys, or swing music intro. What we had to offer Igal Perelman (then at Voxer, now Director of Product at Intuit) was some rather tasteful exposed brick to stand in front of.

This actually fit his style.

Igal’s talk is about the fundamentals: Acquisition, Activation, and Retention (with some nods to Revenue and Referral). Even if you already AARRR a pirate, it never hurts to revisit these basics. Without them, you’re lost.

The talk is elemental. Its deck is a series of black slides with a couple of bullets on each. Igal occasionally refers to the deck, but there’s no script for him to cheat off of. Every point he needs to make is outlined and he fills in whatever else.

Igal’s able to do this in part because he clearly understands these engagement metrics inside out. He says only as much as he needs, but remains a completely natural human. The result is that he’s always formal while still being captivating.

The lesson here is, leave some negative space. If you choke your presentation with a script, with stuffy slides, with use cases you’re not ready to dig into, you’ll be a mess.

There was a time at Mixpanel when all we needed to be engaged was exposed brick with Igal’s sparse fluency to guide us. In fact, even with the bells and whistles, that’s still at the heart of most great product talks.

We still think simplicity works.

And in conclusion

You’ll figure it out. You know, be charming and stuff.

Just don’t be this guy.

Get the latest from Mixpanel
This field is required.