AI Doesn't Fix Bad Inputs. It Just Makes Them Louder.

Welcome to the Marketing MadMen. Nick Constantino here, and after quite a hiatus we are back. Let me introduce Shamir DuRusso of Smart Panda Labs. Shamir, how you doing, bud?

I'm doing well. Thank you so much. Thanks for having me on, Nick.

Let's get everyone up to speed on why you're qualified to have this conversation. Talk about your journey to where you are, and about Smart Panda Labs.

I've been in marketing for about twenty-five years. I started working with Southwest Airlines, Disney, Universal, Marriott, some major brands, and watched the internet take off as a core, fundamental part of business and marketing, killed a number of very large companies we grew up with, and let other companies thrive in that environment. What I saw was this intersection digital marketing created between traditional marketing, product and IT, and how many organizations struggled to get those three things aligned to actually collaborate.

Give me a timeframe on when these epiphanies were happening.

Early to mid-2000s is when the internet really took off. By the late 2000s I saw the conflict between those three groups really start to come to a head, and by the early 2010s I said to myself, I've worked through this, I've helped organizations work through this, let me go off on my own and branch out. I started Smart Panda Labs as a full-service organization, but our real strength was never the advertising and driving traffic. It was what people need to do with that traffic once it arrives, and how complicated that gets.

The people doing the advertising rarely speak with the people doing the implementation, and you end up with a chief marketing officer thrown into the middle who has no experience doing either. What I've learned is you need clean inputs before you can trust any output. If I'm a jewelry company and my data isn't going into the system clean, with the right categories and the right matching, whatever comes out the other end is always going to be subpar. Google won't reward you through their purchase feed if it doesn't understand the language. So talk about this Bermuda Triangle you keep referring to.

It's marketing, product and technology not even thinking of themselves as a triangle that has to collaborate toward the same goal. What we see over and over is a lack of appreciation and communication between these groups, and one treating the other as an order-taker. Marketing goes to IT and says, do this. IT asks questions marketing can't answer. Marketing says, you figure it out. IT builds something, brings it back, and marketing says, this isn't what we asked for. It is what you asked for. After that, six or seven figures get thrown down the toilet. Then marketing goes and buys a tool and assumes the tool alone will fix it.

Give some examples, not client-specific, of where you're either completely rethinking how someone approaches this, or handing them a new level of insight.

Marketers often think they know how they want a system set up. With a content management system they'll say, I want to change this and this and this, and because of the lack of understanding and collaboration, they get it, but it ends up taking ten times longer than it should to get their own work done. Something as simple as keeping content up to date on the website becomes ten times harder because they insisted on their own version of it instead of collaborating and using the expertise on the other side of the table.

So let's start getting into practical tips. Talk about traffic conversion and post-click. I'll say I use pixel-based, strong conversion basics rather than impression-based measurement, but talk about what you've seen work.

A lot of attention goes to Google and Meta, and they'd have you believe you spend money on their platforms, people come, and you're done. It's not that simple. Someone clicks the ad out of curiosity, and now they're on the landing page and have to make a decision, and you have to give them enough information to make it. It's no different from a physical store: find the product, choose it, get it in the cart, check out without waiting too long. If something's distracting or a little too hard, people bail. One quick tip: make sure the image and message on the ad is echoed on the page they land on, so there's no uncertainty about whether they're in the right place. Any uncertainty creates friction, and friction causes people to quit.

There is a workaround nobody wants to invest in: have a stellar reputation for forty years, so people don't have to do as much research because you already have five-star reviews and name recognition. Nike sells shoes because it's Nike. But you can also oversell yourself right past the moment someone was ready to convert.

Exactly, if a person's ready to convert, stop selling, get them to the conversion page, and know what the next action actually is. It doesn't always have to be a sale; sometimes it's a consultation. And you need multiple touch points, because a sixty-five-year-old woman and a twenty-two-year-old guy are going to want to convert completely differently.

Let's talk about MarTech stacks, because I'm going through mine right now and one crack in the armor breaks everything at once. Talk about why these stacks are so often under capacity.

Marketers fall into the fallacy that the stack will solve all their problems, that it'll be set up in a week and they'll push a button and everything magically works. It's going to take time to establish the structure and framework you need to get the most out of it. People get confused and overwhelmed, it's not properly put in place, so they end up using maybe twenty percent of the tool's overall power, and instead of learning to use more of it, they solve the resulting problems by buying yet another tech stack, so you end up with five tools all running at fifteen or twenty percent, none of them accomplishing what you actually want because nobody defined the goal first.

One thing Amazon, Google and the big players do really well is a culture of experimentation. It's never finished, and you have to build a structure that shifts with the market instead of expecting to arrive somewhere and be done.

That's exactly it, and you have to put the steps in the right order. Skip one and you've screwed yourself, and now you have to move backwards before you can move forward again.

Let's end with two or three tips. Say you're a fifty-million-dollar business trying to decide whether to stay where you are or scale to two hundred or three hundred million. What should that person be asking about their marketing and their stack?

The first thing they should ask is how they're impacting each part of the customer's journey, not just acquiring them, but helping them decide, helping them purchase, helping them appreciate the value of the purchase, retaining them, and empowering them to evangelize the business. Think about that entire lifecycle and what touchpoints keep people moving through it.

If you think you're getting from fifty to two hundred million by only acquiring new clients, throw the business in the garbage. You'll get the two hundred million, but you'll have said goodbye to the ethos of what you built to get it. You don't scale that way without evangelists and upsell opportunities; otherwise it's just churn and burn and a race to the bottom on price.

You're trying to build an engine that's already running and building momentum, and that doesn't happen if you're only feeding in brand-new people. The second piece is thinking about that lifecycle and asking what mechanisms you have in place to hear what customers are saying, what they're confused about, what they like and don't like, and feeding that back into how you tweak and experiment. Build, measure, learn, that feedback loop has to run through the whole lifecycle, because the market is constantly shifting under you.

That's such a good point, because we rely on reviews so much. We're five-star reviewed with four thousand reviews across the stores, and you just don't beat that, but a one-star review with a closed-loop response matters more than another five-star review. Nobody expects perfection. It's the three-star reviews, the mundane middle, that worry me, because everyone's so obsessed with five stars they forget the one-stars and how you react to them are what actually matters.

The company is Smart Panda Labs. You can find us on our website, and you can find me, Shamir DuRusso, on LinkedIn and Instagram. I guarantee there won't be another Shamir on there. DM me, let's just talk about marketing.