The AI Playbook, with Eric Siegel

The AI Playbook

Episode 199

Today’s topic is AI and ML, and though you may think this doesn’t concern marketing, we need to acknowledge how it’ll shift things.

Up to now, marketing was done on the premise that for a given audience shown a message, some  average percentage, would act on it. With AI, we’re now able to look at individual audience members and predict how each of them would act upon a message, and at the opportune moment we could have the message show up to each one of them. Goodbye analyzing what happened with crude audience averages, Hello to using detailed data to predict what’s likely to happen. 

With AI holding such promise, why don’t more companies hand things over to AI? I had thought it’s held up by a lack of technical people who know how to do this, but our guest says we’ve had enough technical expertise – He himself was previously one of those data people, and his expertise wasn’t enough to do the job.  He says AI initiatives are held back by those running business functions like marketing who haven’t made the business case and collaborated with the data people to implement this. 

My guest is a leading consultant and former Columbia University and UVA Darden professor. He is the founder of the long-running Machine Learning Week conference series, a frequent keynote speaker, and author of the bestselling Predictive Analytics: The Power to Predict Who Will Click, Buy, Lie, or Die. In 2023 he authored “The AI playbook”

Let’s talk to Eric Siegel.

Timestamps/Chapters:

0:00:00 Intro
00:01:37 Welcome Eric Siegel
00:01:56 Barrier we face isn’t technical know-how
00:06:05 Despite a strong start – AI’s been slow to spread
00:11:17 Process a business needs to implement ML
00:27:41 building a custom algorithm
00:29:45 PSA
00:52:32 The human-side of the switchover
00:54:03 Contacting Eric

People, products or concepts mentioned in the show:

Eric speaks at: Generative AI Applications Summit and at Machine Learning Week

Reviews of The AI Playbook and book’s site

Eric works at Gooder.ai

Geek Professor Drops Rap Video, Tries to Dance

The AI Playbook | Eric Siegel, author | bizML

Clayton Christiansen

Malcolm Gladwell

 

 



Listen to episode

AI playbook diagram

Ecosystem-Led Growth, with Robert Moore

Bob Moore ecosystem-led growth

Episode 198

A pretty widely held view in the world of B2B products is that sales has gotten harder, not easier. It’s not that buyers aren’t buying. By definition, buying is something they do. But in the example of software, some sales reps won’t even know they were being evaluated, let alone passed up for a rival’s product. Only the winning vendor knows that that account uses them for that specific function in their technology stack. All other companies are in the dark.  

 

But are they really? Another way to look at this is that every vendor has information that could be valuable to others. You can find many buyers stacks with products having some overlap but that largely complement each other. As proof, note that lots of these products even integrate with each other because of buyer demand. 

 

Should vendors consider collaborating with vendors they compete against? Aren’t we supposed to hate the competition?

 

We don’t have to. A famous example of that was Apple’s announcement in 1997 of the deal it struck with Microsoft. Steve Jobs defended the deal saying  “If we want to move forward…we have to let go of this notion that for Apple to win, Microsoft has to lose.”

 

Zooming to today’s reality, It makes a lot of sense for vendors to collaborate as part of an Ecosystem. By pooling their data together with their indirect competitors, they can see internal buying patterns. Those vendors who hitch their data wagons together get around the ‘nobody talks to our sales rep’ problem, because one of their partners already has the info that rep needs. Using this intel helps them come first in the race for their product to be selected to go in the buyer’s stack. 

 

Our guest today got a Science & Engineering degree from Princeton University and after a stint in the investment world, he dove into co-founding startups. The first was business intelligence platform RJMetrics and the other was cloud data pipeline company Stitch, both of which he saw through to successful exits. 

 

His latest role is as Co-Founder of a platform that safely shares data among companies for this kind of partner-based selling.

 

Outside of work, He is a Trustee for one of America’s top centers of science education and development And an improv comedy performer, in a  team that has performed over 100 shows together.

 

This husband, father of two, is very proud to call Philadelphia home. Let’s head there now to meet Bob Moore.

 

Timestamps / Chapters

0:00:00 Intro

00:03:46 Bob’s thesis on how sales is broken

00:11:21 Ecosystems are cause for hope

00:26:13 PSA

00:26:53 Revamping corporate partner practices

00:31:38 Pooling together data

00:55:06 Contacting Bob

People/Products/Concepts Mentioned in Show

Ecosystem-Led Growth book

Bob on X

Bob on LinkedIn 

Bob is formerly Co-founder of Stitch Data

Bob is currently CEO at Crossbeam

Metcalfe’s Law

 

 

 



Listen to episode

Partnering on Customer Acquisition, with John Wright

John Wright Partnering on Customer Acquisition

Episode 197

Today, we are going to talk about how those of us who sell things find new buyers once we’ve exhausted our own audiences. We involve partners, and we can do this in a few ways. These partners may have high-traffic sites or be social media influencers. We are trying to use someone else’s channel to reach their audience, hoping they will buy from us.

Alternatively, we might be the ones who are influential or have a large audience that brands want to reach, so they pay us to be their marketing channel. The name for teaming up like this is affiliate marketing.

Today’s guest came to affiliate marketing through dabbling in online gambling. He watched the incentives sites put out to attract players, and then in 2010, he created a website that reviewed gambling affiliate programs called Gaming Affiliates Guide. This site’s traffic led him to become, you guessed it, an affiliate. Over time, he managed several gambling affiliate sites.

As you progress in this field, you always hit a ceiling with this marketing channel. No matter whether you’re the one needing traffic and paying for it, or the one who has traffic and is turning it into money, everyone gets a headache tracking it. As our guest was deeply involved at this point, getting paid to manage affiliate sites, he saw numerous problems in this industry and saw a way to solve them.

There were already applications that reported affiliate activity, but he saw these technologies’ shortcomings. With his engineering degree from the University of Toronto, which had taught him how to develop things, he joined up with partners to create a SaaS tool of their own: StatsDrone.

Having scratched an itch he experienced earlier in his career, he now heads a team whose tool addresses affiliate challenges.

Let’s go to Montreal and hear from John Wright.

 

 

Chapter Timestamps:

0:00:00 Intro

00:03:35 Welcome John Wright

00:06:57 Difficulty with Affiliate tracking

00:11:27 Postbacks and tracking methods

00:18:48 tracking dynamic variables

00:23:14 PSA

00:23:54 Tracking affiliate dollars

00:42:13 Contacting John

People, Products and Concepts mentioned in Show:

statsdrone.com

[email protected]

StatsDrone on Instagram

 



Listen to episode

Mastering Video Ads on Social, with Nikki Lindgren

Mastering video ads for social

Episode 196

 

There’s something we take for granted these days, something that wasn’t even possible a short while ago. Let’s go back to 2008, to the first iPhone, the 3G. What you could send & receive with one, if you could afford the data plan, was restricted to voice, text & small images. That’s because at the time, the cellular networks could transmit at around a third of a Megabyte per second, which went up to 2Mb/second when 3G was fully available. Then LTE/4G started becoming available in North America, reaching 97 percent by 2013. With those data speeds, you could watch brief standard definition videos, and social networks like Instagram & Snapchat began letting you record and send short clips. By the late twenty teens, advanced 4G infrastructure was fast enough, from 12 to 80 MPS, for people to watch 4K videos on their devices, bringing platforms like TikTok along with it. Now with 5G out, lag-free high-def video is available almost everywhere. And if you are a marketer trying to reach consumers, it means that  video must be part of the mix. 

 

There are still quirks to these platforms that we need to figure out. Some of their ad units include ecommerce options for selling products while the ad’s in front of them. More broad that this, it’s hard to know how these platforms will react to videos you post. They know so much about a user’s privacy, it’s raised issues of which country that data’s shared with. Clearly, this calls for an expert’s help.  

 

Our guest graduated from San Francisco State University and FIDM with a business degree and started working in-house at consumer eCommerce brands, running their digital marketing programs. After helping brands in every category from skincare & cosmetics to Books to jewelry, she built her own agency team to do this, Pennock, which is named after the rural Minnesota town where her family are from. 

Let’s go to  Northern California where she lives with her husband Tyler and three kids, to talk to Nikki Lindgren.

 

Chapter Timestamps:

00:00:00 – Intro

00:03:12 – Welcome Nikki

00:09:05 – Video on platforms like TikTok

00:23:37 – PSA

00:24:26 – Reporting to stakeholders

00:29:59 – Ad campaign optimization

00:35:05 – Contacting Nikki

 

Listen to episode

Analytics – in-house or outsource? with Luke Komiskey

Luke Komiskey

Episode 195

We all want our organization’s decisions to be driven by the numbers. Who wouldn’t want to have at their fingertips analytics that accurately show which course of action will be best.  

But doing this takes analysts, and that doesn’t mean hiring them, it means managing them to function well. It means creating processes for them, Outfitting them with technology. Giving them budgets.It’s hard pulling this off in a small or mid-sized organization, and even leaders of large organizations must exercise care when creating this. 

But there’s no set-in-stone law that says a data team must be in-house. Another model, managed services works well for IT and it can be used to give companies access to analysts so they can still be data-driven. 

We’re going to explore the outsourced analytics model with today’s guest. 

Throughout his career, he has worked at the intersection of data, business, and strategy consulting. He earned his Bachelor’s Degree from the University of Wisconsin-Eau Claire.

Following graduation, he joined Cargill as a Data Engineer from June 2011 to November 2013. He went on to serve as the Analytics Lead at Slalom from December 2013 to February 2016, where he claims to have been Minneapolis’ first Analytics Hire. 

In 2017, he co-founded DataDrive, a managed service provider specializing in analytics, alongside fellow data enthusiasts.

Let’s talk with Luke Komiskey.

Chapter Timestamps

00:00:00 – Intro

00:02:14 – Welcome Luke

00:17:38 – PSA

00:18:16 – Calculating value of having good data

00:49:29 – The MSP model

00:49:59 – Where to contact Luke

People/Products/Concepts Mentioned in Show

 

Luke on linkedin

Luke’s company is Datadrive

Jerry Macguire



You may want to hear related episodes:

https://funnelreboot.com/episode-153-boosting-ga4-with-bigquery-with-johan-van-de-werken/

https://funnelreboot.com/episode-151-analytics-worth-the-investment-with-martin-mcgarry/

https://funnelreboot.com/episode-138-how-data-pipes-operationalize-analysis-with-noah-learner/

 



Listen to episode