by Elad Gil
Mariam: Well, I think you have to look at your growth curve, your year-over-year growth rates in your core, and try to project out when you think there will be some inevitable decline in those growth rates. Then figure out how many years will it take to get the second thing that you need to layer into your growth cake ready to actually take its place. And you can model that out and try to predict that.
I know it sounds like I’m making something very complicated sound easy, but we try our best to model that out. Some companies have to do it earlier than others. It just depends on the strength and the legs of the core business, the first business, and how breakable versus non-breakable its customer acquisition model is in the beginning. How self-powered, how self-generating that is. So it’s really about trying to predict the core and what’s going to happen to the core.
The problem with this, of course, is that the second business typically has to get started pretty early. It could be a business that’s growing 500% year over year, but if it’s such a tiny part of the mix, the problem is the mix. It still blends down very low if you’ve got a core that’s growing too slowly.
So in general, it often takes years for companies to really achieve traction with new ideas and new businesses. It does point toward doing what Google did, which was start a little earlier than you think. So that’s the conclusion that I’ve come to so far.
I’m really glad I started the art business, because it’s a very high-growth business, and it’s enough of the mix right now to have a significant effect on the blended growth rate of the company. But we’re going into our fourth year of the art business. We did a little baby launch in 2012, and we started Minted in 2008, to give you a sense of how many years it took before we decided to diversify. It took a while because we were concerned about the scaling problem.
The complexity or difficulty you add to your company significantly affects scaling, much more than just the number of people you hire. The number of people thing is easy compared to the other issue to me.
Elad: How did you think about the trade-offs between, for example, internationalizing more aggressively versus entering this other market?
Mariam: This sounds very simple, but in our particular case I was concerned that the photo-card market was not the same internationally, culturally, in other places as it was in the U.S. That’s one issue.
The other issue was that I really wanted to make sure that Minted’s core was about design, not about stationery. And so what we needed to do was in our core market, the U.S., make sure that our positioning was starting to shift to where I wanted it to get to, which was not being pigeonholed into stationery, but being thought of as just a broader design marketplace.
I knew if I went too long or too far in stationery, I would have the Zappos problem. I like them a lot, but I always know Zappos as shoes, even though they’ve been trying to sell accessories and other things for a very long time. I just can’t get them mentally out of the shoebox, if you will. I didn’t want that to happen to Minted. So I had to move a little earlier in the expansion of what the brand stood for.
And then the other thing is, I felt that the complexity of going overseas was pretty massive, actually. The trouble that it would place on the company’s executives to have to split their time in a very, very taxing way, involving a lot of travel and a lot of hours, I just felt was going to really break us. It would actually be a little easier to start a different business in the U.S. than to take the other business overseas.
Elad: When you decided to start that new business, how did you think about hiring an initial team? Was it just transferring over people who were internal? Did you hire anybody external?
Mariam: From the beginning, we tried to put very little money into the test. It was very much an MVP. So I integrated it into people’s existing jobs. The merchants, the people who ran the competitions, they all just started launching art challenges. And then the supply chain people started trying to figure out how to print and frame art. They clamped it onto their day jobs, to try to see whether this test would work.
And then when the test began to work, the first hire we made was a woman early in her career, right out of Stanford. And she basically was the entrepreneurial “I’ll do it all” kind of person. Because we still didn’t have enough scale yet to throw a huge amount of loss into this business. It’s kind of crazy: I think we built a couple million dollars of revenue with that one great hire and by leveraging a lot of other people. Then it got to the point where it became clear that the business needed its own set of marketing initiatives and strategies, and we hired a director of marketing for art. It’s a matrix organization; it’s not a GM organization.
We are thinking about experimenting now with a GM organization. It’s just that we were warned that we needed to be careful about a GM structure because it carries the risk of creating politics. That it would silo the brand away from the main Minted brand, and that it wouldn’t hang together.
“There’s another dimension which really significantly affects scaling, and that’s the differentiation or complexity of the business.”
—Mariam Naficy
What we’ve found is, if you’re going to think about a GM, look for a GM who leans toward naturally solving problems together, who’s happy to collaborate, versus the personality who wants to really run their own island—that makes things much harder when you’re running a business where there’s supposed to be this integrated customer presentation, a holistic presentation of the brand and the company. It’s much easier when you have people who are very collaborative at the top of those internal businesses.
It’s a little bit of a personality test actually. I’ve had to develop interview questions to try to suss out who would be good as a GM in this kind of model. But that’s one of the really, really important attributes: whether they’re truly willing to make collaborative decisions about the business with the CEO and with other GMs.
What I asked this person was, “What would you want from me as your manager? What kind of interactions do you want?” You can tell from the answer whether the person really wants to have anything to do with you or thinks that you have something to add.
The person in the interview kept saying things like, “I will communicate just to make sure that you have absolutely no negative surprises.” I don’t want a relationship where we only communicate because you’re worried about me having a negative surprise. First of all, that’s not our culture, to punish people for a negative surprise. That’s a big company kind of answer. So I was a little wary about the answer that I got.
“How do you communicate with your manager?” is another question. “How do you like to communicate? Formally? Informally? How frequently do you communicate with the person that you currently report to? What’s the style, and how do you get in touch with this person?”
Elad: I actually think that’s a really good general executive hiring question: “What would you want from me?”
Mariam: Just to understand this person’s communication style is really critical. But then you can get all this other stuff out of it. “What do you want from your peers?” is another one. Either question will suggest whether the person really thinks their peers or their manager has any value to add at all, or is simply there to exist. So you can get a lot out of asking them about what they plan to seek.
The hunt for the GM is difficult. Turns out it’s very hard to find good general managers because they have to be very good generalists. A lot of times they want to run their own companies, or they’re already running a company. They may not start a company, but they may want to run a company. And when you have a couple divisions, you want all the GMs to be of equal caliber. Also, depending on the way the rest of the company’s structured, the core competency of each GM will differ.
For example, let’s say I have one GM who’s very merchandising oriented, because the person came out of merchandising. And another one who’s very strong at marketing, because they came out of marketing. And another
one who’s strong at product marketing, because they came out of PM. Creating processes that work equally well across all of them, given their different strengths and weaknesses, is actually really difficult.
I will just say that this idea of running multiple businesses has been among the most complicated parts of scaling. It’s really different for each business, where you are and what your competitive landscape looks like.
Elad: One other aspect of scaling that many companies run into is technical debt. There are certain parts of engineering infrastructure that companies need to rebuild or invent over time. How did you work on the technical debt? Do you do fix-it weeks? Do you have specific projects focused on it?
Mariam: What we do is we figure out our growth road map for the year. Then we take the pay-down projects and we piggyback them along the growth road map. So we’ll choose to do things in the order that the growth road map dictates.
For example, our product details page—it’s the page you land on when you’re looking at specific products. We had let the company launch about six different versions of this page, because a lot of different teams had worked on it. All of a sudden you have all of this repetitive, redundant code. Now we’ve rebuilt it to have one code base, and we can swap in different modules. So that obviously does a lot of things for us. But we did it first because that’s one of the key things that was blocking the growth road map. That’s one example of how to prioritize.
You also need to keep an eye on the productivity of your team. Once you see engineers starting to really slow down and be afraid to touch certain parts of the code, I think you, as a responsible person, have to create a balanced scorecard. Even if it’s not directly linked to a revenue project, you can always come up with a measurable impact, hopefully, for a sound strategy. So as long as there’s a metric involved, I think that we feel comfortable arranging the road map and figuring out how we can pay the tech debt down.
Prioritizing it, for us, is mostly determined by the growth road map, but then sometimes it’s about productivity of people. And sometimes it’s about pure happiness. If there’s a tipping point where people are really unhappy about something, obviously we do take that into consideration.
The capital environment—whether capital is cheap or expensive at that time—also vastly, incredibly influences that scorecard. If capital is super cheap, you can get tons of capital to last you for a long time. And so you can think very long term. You can prioritize long-term-payoff projects. If it’s very expensive, you have to start pulling your horizon in and looking for shorter-term returns, which might mean lower beta projects, things that pay off in the year that you do them. That’s a very tight screen for what you might pursue. Paying it off within twelve months is a tall order, I should say. Not a lot of projects qualify.
I like to use the balanced scorecard. I do also, in general, like to prioritize revenue growth. That’s probably the top one for me, but then there are other important things like brand, customer satisfaction, community satisfaction, etc. on our scorecard. But we don’t like to do massive tech debt pay-down projects that have no connection to the business plan of any kind.
Elad: How do you create that culture? Because I think that sometimes there’s that natural tension between engineering and the business side—engineering wants to work on projects that are either interesting or that address pieces of infrastructure. But sometimes those pieces of infrastructure, to your point, aren’t really necessary.
Mariam: We try very hard to have engineering and product meet to discuss ranking and to really look at the company’s goals together. Essentially at the very top level, people will leave if they don’t believe in being part of a business, and they don’t believe in the company’s goals. They will self-select out. So hopefully in the interviewing process, we’ve explained, for example, that it’s an e-commerce business, it’s exciting for these reasons, we’re building a design community and a marketplace. And one of our key metrics for success is actually size of revenue. Hopefully on the inbound, you’re telling people what you value, and then people are selecting in based on that. I think the recruiting process is critical to solving that problem.
But in general, we communicate many, many, many times what the overall strategy of the company is, and the corporate goals. Finance has given classes in ROI calculations to engineers. We have a lot of finance classes, essentially, that we’ve organized for the engineers to understand how to commercially evaluate these things.
“It’s very hard to find good general managers because they have to be very good generalists. A lot of times they want to run their own companies, or they’re already running a company.”
—Mariam Naficy
Elad: That’s a great idea; I think more companies should do that. I’m curious how you came up with that.
Mariam: I think because I kept talking about ROI and related areas so much. Finally the PM/engineering leaders said, “I think we need to tell people what these things mean.” The Head of PM actually came up with the idea of just having our VP of Finance give these classes.
The funny thing is, people seem very interested in it. A lot of the engineers come up to me after all-hands meetings and ask me questions about the stock market and stocks that they follow. Not that I’m an expert at all. But there’s just a lot of interest in, you know, “Will Minted go public?” or, “I’m reading in TechCrunch that these things are happening. Why are they happening?” So there’s actually a fair amount of interest among our employees.
I think we also have a language inside the company. Let’s say you’re logging into some Google Doc that people are using to rank projects, you’re going to see all these financial terms, like “ROI” or “net present value of the revenue,” because we use that language a lot. And so I think that helps. That’s another thing that’s just really embedded in our culture, so deeply that we don’t even notice it. I’m a former investment banker, and my cofounder, Melissa, came out of the finance department at eBay. So we’re both numbers-oriented.
Elad: How do you think about seniority of the team versus scaling or functional debt accruing?
Mariam: This is really important, and I think what it says is that perhaps the first place to really scale, and I mean scale to senior level, is engineering. You should never start with the cheap solution of building a whole team of young engineers. Whereas there are other functions where you can potentially get away with it.
Elad: What would be an example of a function like that?
Mariam: Things that are so easy to see on the surface that it’s clear as day whether the result is good or not. So, for example, design. I think you can actually start with very young, emerging design talent, because you can see right on face value whether the work is good or not. Whereas when it comes to things which are more hidden and harder for the CEO to see clearly, you might want to hire more senior people, like in engineering—the architecture, and what’s happening with the systems of the company.
Elad: That’s a great framework for senior versus junior hires. It’s okay to hire junior people for areas where it’s clearly visible to the CEO if it’s working. And any area where it’s hard to understand things quickly and easily, then you should always have senior hires. I think you should call that the Naficy framework.
Mariam: That’s how I operate. So I’ve hired very young people and junior people in design areas and merchandising, where I can see everything.
The only place where I think engineers might get into trouble is on the marketing/finance side. I see that a lot. There’s this very important issue of getting your business model straight. And just as businesspeople have trouble figuring out who’s a great engineer, I think sometimes engineers have a really hard time figuring out who’s a good business person. It’s just this crazy divide unfortunately. There are not that many people who really understand both the financial side and the engineering side.
I’ve seen engineering-oriented founders struggle a little bit trying to figure out, “Do I
trust this person or not in engineering my financial structure, or my marketing/financial structure, my customer acquisition stuff?” And some of that can go really wrong, really quickly. You can go really wrong spending money on customer acquisition if you don’t know what the right metrics are. So that’s one of those things that you need to hire for carefully.
It doesn’t necessarily mean you need a senior person, but you need a really, really good person. And that person is probably not the cheapest hire.
“If capital is super cheap, you can get tons of capital to last you for a long time. And so you can think very long term. You can prioritize long-term-payoff projects. If it’s very expensive, you have to start pulling your horizon in and looking for shorter-term returns.”
—Mariam Naficy
Elad: The one other thing I’d add to your framework is that there are certain specialist roles that are existential to your company, and that’s also where you want to hire senior people. So an example, back when you still had data centers, the person who was running your data center build-out had to have done it before. Or in the case of Color Genomics, we needed somebody senior who’d run a clinical lab to run our clinical lab.
Mariam: Yeah, exactly. You guys have some serious governmental and regulatory issues that have to be contended with. I totally agree on that. Legal is another area where I really don’t think people should go cheap and junior. The legal stuff can get really bad if you don’t get someone experienced. I mean, I think that’s probably obvious, hopefully.
