AI and … Pizza?

The Italians first invented pizza roughly 1,000 years ago. We can only assume the first developer meeting was scheduled for ten minutes later. Otherwise, whatever did they need the pizza for?

Now, seriously, it’s fair of you to ask what in the world pizza has to do with AI and digital strategy. A lot more than you might think. Here’s why.

Pizza was one step into the future, a dish that would last a thousand years. AI is another step into the future. Just the far future. In fact, it’s a reality right now. One of my favorite quotes says “The future is already here. It’s just not evenly distributed.” It took centuries for the chewy, wonderful goodness of pizza to make its way around the world. It will take time before AI is “everywhere.” But don’t think it’s not around just because you don’t see it every day.

Google, YouTube, and Facebook use AI in the algorithms that determine which websites, videos, and shares you see on their respective platforms. The Associated Press, Washington Post, and other media outlets routinely use AI to develop content and create rough drafts — and not so rough drafts — of articles for publication. And one of these days, you can bet someone’s going to teach an AI to develop the world’s perfect pizza.

The point is that it’s time for you to start thinking about how you plan to use AI to improve your business. And the best way to do that is to order a couple of pizzas.

No. Seriously.

Jeff Bezos at Amazon popularized the idea that to get something done effectively and efficiently, think in terms of “one pizza teams” and “two pizza teams.” By which he meant that the best teams — where best is defined by quick and effective — were teams that you could feed with no more than two pizzas. Any more than that and you’ve got too much overhead, too much cross-talk to truly be effective. There’s a bunch of well-understood math that explains why two pizza teams make sense. (BTW, Fred Brooks’ classic project management text, “The Mythical Man-Month: Essays on Software Engineering,” said the same thing almost 45 years ago. He just didn’t use the terms “one pizza team” and two pizza team.” I suspect that Brooks was probably more of a chateaubriand guy than a pizza connoisseur).

The reason some companies are struggling to figure out where AI fits into their businesses is that they either have too few people working on the problem or — far more likely — too many.

The right way to figure out how AI is going to work for your business is to assign a small group, one that you can feed with a single pizza (or two, tops), to investigate business problems that:

  1. Have clearly defined outcomes. You know what you want in terms of results. And…
  2. Currently flummox your organization. Even if you know what you want to accomplish, the issue to date has consistently resisted efforts to automate and improve.

There’s an old joke that claims a camel is nothing more than a horse designed by a committee. Want a better horse? Kill the committee. Focus on the folks who add value and ditch the rest.

If the puzzle you’re trying to solve requires a group larger than a two pizza team, break it into smaller pieces — kind of like “slices” — and assign those to your small, nimble team. When successful companies talk about “agile,” “lean,” or associated methodologies, that’s what they’re doing too.

Artificial intelligence isn’t some magic pixie dust you sprinkle onto existing initiatives in hopes that it will make some spectacular difference. It takes work. That work can be at enabled by focusing your team’s efforts in an effective direction and reducing the friction that frequently limits success. And, of course, fueled by a slice of pepperoni, mushroom, or plain ol’ cheese pizza.

So grab a pizza. Or two. But no more. Then round up a few folks at your company who like pizza and like learning to get started on putting AI to work for your future.

Happy Pizza Day, everyone!

  • Footnote 1. Yes, I’m aware pizza had a number of precursors like flatbreads that probably existed for thousands of years before the date I’m citing above. I’m using Wikipedia’s dating. Go fight with them if that matters to you.
  • Footnote to Footnote 1. Also, the stuff we think of as “modern” pizza probably only dates back to the 1800’s before emigrating to New York and New Jersey where we perfected it.  [Editor’s note: We also think Chicago deep-dish is pretty delicious.]
  • Footnote 2.Though I’d argue that the folks at Razza in Jersey City already have developed the world’s perfect pizza. Fight me.
  • Footnote 3. Just please, dear God, no Hawaiian. Yuck. [Editor’s note: Our correspondent could not be more wrong on this one. Who doesn’t like pineapple on pizza?]

About Tim Peter

SearchChat Podcast: How Facebook Got Sent to App Jail

Facebook is having a terrible week. After experiencing a barrage of trouble over the last few months, they’ve finally crossed a line Apple won’t tolerate. They made available an app that gave themselves a scary amount of access to your device. It’s opt-in, but Facebook seems aware that it’s invading privacy — and appears to be preying on young people.

How well do people understand how you’re using their data? 


We also discuss the top trends people are talking about in 2019. After some keyword analysis and the input of sites like BiznologyCMO and more,  we can tell you all the most important digital marketing trends to watch. The biggest name will be no shock: Artificial Intelligence.

But do executives really know how to implement AI technology in a way that works, to create a seamless learning experience? The secret is starting small, with just what you know.

0:00 Intro

2:05 Facebook’s in App Jail

14:45 What are Top Trends pages saying?

17:40 How can executives get started with machine learning?

24:15 Seamless customer experience

27:00 Outro

SearchChat is available on

Originally published on Biznology

About Tim Peter

Why AI Has Come a Long Way Since HAL in 2001

January is a special month in AI history. Because in both the novel and movie 2001: A Space Odyssey, January 12 is when the HAL 9000 sentient computer — (spoiler alert!) the story’s antagonistic artificial intelligence — goes live. Depending on whether you date HAL to its “birth” in the film, the novel, or when those media originated, HAL is anywhere between 22 years to 51 years old now (For trivia buffs, of which I’m one: The book and film were released in 1968, making HAL’s conception over 50 years ago; if you go by the dates given in the film or the book, respectively, HAL is either 27 or 22 years old). HAL is then placed aboard the Discovery One spacecraft to participate in a journey of, well, discovery to the planet Jupiter.

About Tim Peter

SearchChat Podcast: Can You Personalize Without Creepy Data?

Is the dream of the visitor journey dying? How do we make journeys more functional without using data people don’t want us to have?

Marketers are starting to learn they can’t just orchestrate a visitor journey from start to finish. It’s all about improving the journeys people actually make. They’re complex, not straightforward. Steve and I discuss how visitor journeys are a big data problem. Machine learning allows you not to have to manually deal with that data. It makes Big Data little.

Data can be put to work automatically to make the journey better — and it doesn’t have to be a ton of data. We often start with search data, and it works great since it connects people directly with the thing they want. “Behavioral personalization” means personalization but without all the creepy data. Instead, it’s personalization that customers are asking for. This matters in a post-GDPR world.

Google’s policy is to get right up to the creepy line without crossing it.  Most people don’t know that smart TVs are cheap because they are tracking your data.  How long will models built on creepy data survive?

The three laws of robotics initially were just about making sure robots don’t kill humans. Now we’re thinking much further beyond that — how to create ethical artificial intelligence for business.

Tune in and discover more!

00m 00s — Intro and overview

2m 00s Visitor journeys are changing

7m 05s AI for developing visitor journeys

11m 05s Behavioral personalization

15m 25s Creepy Data

19m 30s 3 Laws of Robotics — how do we create ethical AI?

22m 45s Is it just “legal,” or is it good for customers?

29m 35s Outro

SearchChat is available on

Originally published on Biznology

About Tim Peter

SearchChat Podcast: Ring in the Year by Putting Data to Work

Analytics matter: this is the unavoidable fact of digital marketing, even for those digital marketers that fear it. But are you even measuring the right things? Do you know how to make meaningful improvements?

In this episode of our SearchChat podcast, Steve and I talk about site search, personalization, and big data. In our work in website search, we’ve seen that clicks are a measure of activity, but not necessarily an indicator that something good happened. Did the click lead to a purchase? Did the click answer to a visitor’s question?

First, a brag: Marketing Tech Outlook named SoloSegment to its top 10 marketing analytics solutions. We talk about what we’ve learned and what we now offer our customers. When I first heard about receiving the award, SoloSegment was mostly collecting data. Now, we realized what sets us apart is automating changes using that data.

Our focus for 2019 is on  putting data to work. It’s not an easy task — it means determining if your data is accurate, as well as usable to measure success. 

We discuss personalization, which every marketer wants to jump into. Not everyone is ready.  Do you have the data to identify your audience, what the right content is, and identifying whether it’s working or not?

Tune in and discover more!

00m 00s — Intro and overview

02m 00s — SoloSegment named in top 10 marketing analytics solutions

5m 20s — Why measurements like clicks fail

9m 25s — Can you use your data to power success?

15m 15s — Why your B2B content marketing isn’t ready for personalization

20m 45s — How to think about Google Discover

28m 02s — Subscription links and outro

SearchChat is available on

Check us out on FacebookTwitter, or email info@solosegment.com.

Originally posted on Biznology

About Tim Peter

SearchChat Podcast: AI Goes Back to the Basics

We at SeachChat frequently talk about how AI and site search produce value for your site. But let’s break that down for a minute. What this is all about at the end of the day is customer experience.

When a prospective customer arrives on your site: are you helping them? Are you answering their question? What value might you be creating — for them, and for yourself?

Steve and I focus on some of the most important ways to fix your site search improvement program. It might not sound like the most glamorous solution, but it’s the best way to ensure you can capitalize on site search insights. Site search offers some valuable information: what can you learn about a visitor and their intent.

As I wrote recently, site search is your company’s best salesperson. When powered by AI, your site search learns about your prospective customers and can tailor results to guide them. Machine learning lets site search deliver results that drive sales. If a salesperson was performing as poorly as your site search, would you even keep them around?

00m 00s — Intro and overview

02m 20s — Site search insights on Search Engine Land

13m 00s — Site search value and site search as your best salesperson

18m 50s — Developing a strong site search improvement program

23m 16s — AI and its connection to search

32m 30s — Customer experience

33m 23s — Subscription links and outro

SearchChat is now on

Check us out on FacebookTwitter, or email info@solosegment.com.

Originally post on Biznology

About Tim Peter

SearchChat Podcast: Budget Season Survival Guide

Not enough marketers take advantage of the other kind of search — the one on your own website. Few companies budget for it, while budgeting for content without a second thought. We’ve talked about the cost of value before. But when they search, can visitors even find the content they need on your site?

Steve and I are excited to introduce a new podcast, exploring the topics we are fascinated by: AI, search, and content. Site search is part of a customer journey. When you optimize your site search with automation, visitors can find your content and continue on their journey.

Today we cover the Budget Season problems: proving why site search matters, what makes for good analytics, and how much budget you need to make your search better.

00m 00s – Intro and overview

01m 17s – Start of discussion with Steve

07m 04s – Do clicks mean success?

11m 44s – What do we mean by upstream/downstream traffic to/from search?

13m 12s – Why it matters that Google exited the site search market

14m 58s – How much budget is enough to make your site search better?

17m 27s – How can you get started on improving site search?

SearchChat is now on

Check us out on Facebook, Twitter, or email info@solosegment.com.

About Tim Peter

Warning: You’re ignoring your company’s best salesperson

Here’s a scenario for you: imagine you have an amazing salesperson who develops a deep connection with customers, beginning with their very first interaction. Even better, these prospects share their deepest concerns, telling your salesperson everything you’d want to know about how to help them — and how you can sell them what they need.

But you ignore everything this salesperson wants to share with you about what they’ve learned. You simply say, “Nah, I’m not interested in providing a better experience for these prospects. I’m not curious about their needs. I don’t care what they’ve told you.” That would be ridiculous, right? And yet, if you’re like most companies, you’re probably doing this every single day.

You may have guessed that your company’s best salesperson is, of course, your website. This brilliant salesperson who knows what matters most to your prospects and leads might still surprise you: website search. That is, the searches customers conduct directly on your site. What customers tell you in those searches will make the difference between successful enterprises and the also-rans.

About Tim Peter