Finding the logic in the illogical

Living in your partner's world of problems: The IMDb online ticketing story

The moment it all made sense…

Sometimes the thing is not the thing

• “Those sales people always have some excuse why it’s not their fault the product not selling.”

• “Sales says if we don’t launch _____ by _____ they’ll miss their numbers."

• “What sales says then ‘need’ to close deals doesn’t seem to provide any value to customers. WTAF?!”

Most people who’ve worked with sales teams have heard or said one of these things. They’re closest to the customer, but in quiet moments folks may view Sales as more the “voice of sales” than an accurate “voice of the customer.”

Sales folks tend to be pretty good at, well … selling… so it’s easy to take them for an unreliable narrator when explaining why your (perceived) hot commodity isn’t flying off the shelves. Though they’ll make up for it with the best stories if you have a drink with them after work.

Today’s post though isn’t about the psychology of sales, nor the relative value of selling through “overcoming objections.” Instead, it’s another twist on my old favorite. The value of being curious long enough to understand why something that makes no sense, may in fact make a lot of sense.

I’d intended this Substack to have of a practical memoir feel to it. This “minor” example actually had an outsized impact on me and I share it often - so it seemed like fair game. It’s totally not just because a few months ago I got ChatGPT to create an alarmingly accurate re-enactment of myself, Jeremi Gorman and Brian Elieson in the (for me) pivotal discussion that I thread through this article.

The feature that made sense but totally didn’t

Things that matter don’t always make sense based on your perception of reality. But they likely make sense to someone’s reality. How it makes sense to them is more important to understand than most of us have been taught. Teaching implicitly or explicitly often centers on convincing others, as opposed to recognizing the moments when we need to convince ourselves of something new.

Today I’ll just share an exampleof such a misunderstanding. Or more accurately, differing assumptions due to incomplete information.

If you’re developing systems long enough eventually you’ll have an moment where is building a feature that seemed not worth the effort¹. “If we don’t match up on features we’ll never get to be in the right room to really sell” might have been the response when you objected. Most likely you just sucked it up and did it if it was small, but grumbled all the way through. I’ll probably write an article someday about how this instinct is super dangerous. As productive dev teams like to build things that get used. The most 100% surefire way to demotivate and decimate your team is to work on things that don’t get used. But that’s for another time.

The perspective on “what features a product is missing” varies based on your seat at the party. As one who wears many hats - but the engineering one being a primary one my view can sometimes include internal capabilities that aren’t visible to users. Thinking back to my tenure at IMDb I believe if you’d asked me what three things that the site was missing I’d have said

• A true database mastering all of the data as opposed to a huge collection of flatfiles².

• Where to watch the content I was reading about. Assuming it wasn’t in theaters, on DVD, or available via Amazon Prime.

• An ability to purchase tickets on the website

The path to address the first two are twisty, fascinating stories for future articles. Today we’re going to talk about the third one → the ability to sell tickets on the website. While I’d have listed it as a “product gap,” I didn’t really believe it was critical to fix. In my labeling taxonomy I’d likely have bucketed is as “interesting” but not ‘important.”

It turns out I was very wrong about that. Thanks to Jeremi Gorman and others we didn’t get stuck in my bad assumptions on reality.

Wait - buying movie tickets didn’t seem important? For a movie website?

IMDb (as now) was considered the most authoritative place to learn about television shows and movies. Advertisers were mostly on board with buying ads on the site for film releases - and shared good feedback on the team’s custom creative solutions that took over the website with the intent of building awareness and directing visitors to the film’s “detail page.” IMDb offered lots of information about where you could then see the film locally (aka - “showtimes”). But what it didn’t have was a way to buy the tickets directly.

Moviegoing was a different world back then. Buying tickets in advance wasn’t the same table stakes it is now. With the possible exception of super big budget films that were likely to sell out during opening weekend. I believe the main shift in valuing online purchasing was driven by theaters offering reserved seats. At least for me, the ability to lock in the spot I wanted to sit, and thus arrive as late as possible to miss the incredibly annoying “pre-show³” was suddenly worth the effort and “convenience fees.”⁴ That wasn’t a common service though at the time of this tale.

There sites for buying movie tickets (largely Fandango) and the data we had indicated that even for big openings a relatively small percentage of tickets were sold online. Thus it didn’t seem a critical lever - and we had a long list of technical and product levers we wanted to pull first. We didn’t have the appetite to try to actually be the point of record for selling tickets, ie; attempting to battle Fandango. If we had thought we could pull that off then this would have been a very different adventure.

Repeated asks from Sales had put an integration with Fandango toward the top of our to-do list. I wouldn’t say the product teams were sold on this. Grumbling from the engineering teams about the execution challenges and relative value have stuck in my memory. It’s challenging to build enthusiasm for building something when one feels their own explanation as to why it’s important is unconvincing. High performance teams want their work to matter, and I was not a great salesman for the relative prioritization of this work. At least not in the initial stages of the journey. Though we progressed nonetheless.

Spoiler alert!

As I’m typing I realize this is going to be a particularly underwhelming narrative. We were working on doing something (ticketing purchasing) - but didn’t feel it would have much impact. Then due to our continued curiosity we finally “understood” how it mattered. And we⁵ finished the technical work and launched the feature. One could argue that the end result was completed unaffected by the “learning” I keep teasing .

That is all true.

Sometimes you’ve just got to get started on things. That shouldn’t stop you from seeking clarity even while underway. It’s never rarely to late to adjust your actions in light of better information. If it turns out you made a not great decision, just move onto trying to make a better one, and then another one.

Where you end up is way more important than where you were heading when you first hit the gas. Due to the cognitive “peak-end” rule, if you start a project not being 100% sure it’s high value for your customers and/or business but you end with clarity it is - almost everyone will remember only the end-point. Similarly, if you start off with clarity but drive off a product-market fit cliff of sorts …. well, I think reasonably that’s what people will remember from the story.

At the risk of saying it more times than I need to, engineering teams really want to build things that get used (or teach them how to do better/faster next time), and from a ROI investment you should feel 1000% aligned with this. Continually confirming everyone’s on the best possible path doesn’t have a start-by or end-by date in my view.

Hopefully your interest is piqued. That’s good - because the true story is almost over finally here. ;-)

Assumptions interrupted (yes, finally the end!)

I’m sure I asked “why is this so important again?” many times. One of those times either the answer was clearer, or I asked better, or maybe I just was paying more attention. That day Jeremi broke through my incomplete understanding of the value of the feature for Sales, and to an extent for advertisers.

Jeremi was a gifted storyteller, and at least pretended that she didn’t mind explaining the same thing more than once. All of her stories were interesting, some extremely funny. Several of which I’m pretty sure I shouldn’t repeat. We haven’t spoken in a while - but I can guarantee she has more of them now than she did then.

Anyway - at some point during the project (or right before we really broke ground) I asked again “I see why people might want to buy tickets online. But given all the things we could build why are you so so adamant this is the ‘most important’ right now?” In hindsight, this is a terrible question - it could come across as potentially accusatory, negative, and it doesn’t soften the blow by starting with “how” or “what.” Even so the answer was spectacularly clear.

At the risk of bolloxing it up the response I heard was

“Well … entertainment advertisers basically have their budgets for ads allocated into three categories.

1. Endemic websites - places that people go to learn about movies.

2. Reach - sites that can deliver tremendous number of (presumably targeted users). Basically Facebook.

3. Websites that sell movie tickets.

Today my team can only access the budget for #1. We’ll never be able to sell IMDb as #2 compared to Facebook. But there are a bunch of dollars in #3 and that changes the game. Just because technically they can buy ads with the money in budget #3.”

In hindsight it was as if the clouds parted and sunshine streamed down from above - I felt simultaneously enlightened and incredibly dumb. My sheepish response, “oh … that makes a lot of sense…” I think I glanced over to Brian Elieson a moment later with what was supposed to be a look of “geez, I wish I’d understood that earlier.” I’m not sure Jeremi clocked any of it as especially noteworthy, as if I was her I’d bet she would have been thinking “these guys keep asking an obvious question and I keep trying to explain it to them. I hope I don’t have to keep doing that.”

FWIW - I’m pretty sure we never asked again.

Aftermath

No one’s memory is especially accurate, and I’m sure there are parts of this tale that are wildly inaccurate. My apologies to anyone involved if that’s the case. What I remember most after the moment of recognition was that

a) there was a pretty darn clear reason for the feature that made perfect sense given the knowledge the sales team had, and

b) I needed to stay curious with good followup questions when things didn’t make sense.

I can’t guarantee you’ll always see that clear reason given enough questions. But I’ve gotten a lot better after that moment in staying curious enough to consistently discover why things make sense to someone, when they don’t make sense to me initially. It’s unlocked a lot of business results over the years since - so thanks for reading this far and being open to hearing the good word.⁶

1. Possibly many, many examples.

2. Yup - no database. Yes - the name of the company used 50% of the letters in the name to refer to a database. But still true at the time. We also had a pretty crazy tech stack - but hey, if it works (sort of) it’s sometimes hard to get the alignment to fix it. I think I’ve alluded to this in past writing, but willing to explore in more detail if there’s interest. [As a footnote to a footnote - arguably a flat file is sort of a database, but I’m not here to argue semantics]

3. Yes - I’m looking at you Regal Cinemas. “Noovie?” - Screw you for exposing me to that incredible mess of advertainment dreck. Note: my hard learned reluctance to arrive early does not apply to Alamo Drafthouse whose focus on customer experience includes curating fun (non advertising) content before each show. Which probably encourages you to buy another beer or whatever.

4. In these politically polarized times, it’s refreshing to see that every once in a while legislation is introduced that I’d think everyone in America can actually rally around. Yes - I’m talking about forcing theaters to publish non-padded film start times.

5. I’d like to acknowledge the awkward reality that “we finished” basically refers to a bunch of work that “I” didn’t do at all - the hard product definition and software development pieces. At best I might have asked some pointed question around what level of uptime, monitoring and response SLA’s to expect from Fandango. You know - “whose phone number do we have to call if this doesn’t work at some point post launch?” I’m not trying to minimize the value of that question (tbh it can get overlooked otherwise). But it’s hardly like I built anything in this story.

6. OK - I’ll cop to it, having that cartoon at the top was a material motivator for me in writing this post. :-)