Enterprise Growth Wisdom From IBM’s Zoltan Szalas
Zoltan Szalas leads growth for IBM’s data and AI portfolio, focused on uncovering growth opportunities, increasing user satisfaction and providing leadership that helps achieve specific revenue objectives.
Previously, Zoltan founded Croissant, the world’s first on-demand shared workspace app. It’s now in 40 countries, has over 10,000 customers and brings in more than $2MM in annual recurring revenue.
The secret to his success? A data-driven approach to growth—and a genuine love for all things data.
Growth at an enterprise organization is much different than it is at a startup. For one, how do you build and manage a growth team? I couldn’t think of a better person to ask than Zoltan. Here a few key insights he shared with me.
Team structure: Depends on your growth stage
Before you find product-market fit, growth is everyone’s job. But once you’ve reached the product-market fit milestone, it’s time to start thinking about how to scale your growth.
There are two primary ways to build out a growth team:
- Bring together a group of existing employees from multiple functions to create a home-grown team.
- Start out of the gate with a more formalized growth practice, staffed with resources who are hired for the purpose.
While Zoltan thinks every company should eventually establish a dedicated growth team, the right time to do so depends on the company’s growth stage. Growth management is still a relatively new role, so it’s often challenging to try to shoehorn it into an existing enterprise organization like IBM.
Most of the time, if you try to force the issue prematurely, the effort will fail. A better approach: ease into things by building a growth team from within—pulling existing people together to tackle the challenge.
With that approach, you’ll still need an experienced growth lead who connects all the parties, manages cross-functional team dynamics and brings ideas to the table. Ultimately, this person’s main jobs are to make sure everyone’s objectives are aligned, and to get consolidated buy-in on growth initiatives.
IBM has developed a formal growth practice around a dedicated team. In their structure, Zoltan manages growth across the entire data and AI portfolio as the growth product manager. But he works with nine individual growth product managers who are each responsible for a product in the portfolio.
Even if you’re not quite ready to start assembling your team, Zoltan recommends taking steps to embed a growth mindset into your organization’s culture sooner than later. The earlier you can help people understand and embrace growth as a functional area, the easier it’ll be to align future teams and coordinate collaboration.
Who to hire: One objective, two kinds of people
Whether you’re sourcing your team internally or externally, understand the two different kinds of people you need to build an effective team: the more senior management role and the more junior “executor” role.
Senior Growth Lead
This person has the experience and ability to handle managing relationships, team dynamics and “politics.” They have exceptional communication skills, and they’re able to apply those skills not only with internal team members and peers, but also in dialog with investors, board members and company executives. This person must have a vision—as well as the ability to articulate that vision in a way that inspires confidence and encourages people to sign on.
Junior growth support
This person is hungry to learn and deeply curious about all things technology and testing. Plus, they’re willing and able to take initiative. This is someone who constantly finds better ways to do things. A natural problem solver who consistently pushes boundaries in the name of higher performance.
When he joined IBM, Zoltan immediately recognized the huge difference between that environment and the startup companies he’d previously worked at. As he explained it: In a startup, you spend 20% of your time managing team dynamics and 80% of your time executing on plans.
And in an enterprise company, the sheer scale of the organization requires that you flip that. There are so many more moving parts and levels in a large operation—making sure everyone’s working together effectively can almost become a full-time job.
Keep this in mind when you’re thinking about who to hire and when to bring them on board.
Growth process frameworks: A necessary guide
When you’re working in an enterprise environment, frameworks are critical because the work you do needs to be both well managed and replicable.
At companies like IBM, much of the success in experimentation and growth comes from improving processes so they work well across platforms and portfolios. This is the “economy of scale” people always talk about.
The framework Zoltan and his team use for growth initiatives at IBM is a five-step process designed around working with large, cross-functional teams.
Step 1: Identify the biggest opportunity
When you’re part of an IBM-sized organization, the best opportunities are the ones you can replicate across different products. Typically, the selection process starts with looking at broad pieces of the funnel, and then narrows down the options by getting more granular.
Part of this process uses another framework that Zoltan refers to as the vertical/horizontal growth framework, a way of assessing an opportunity on net-new successful experiments (on the vertical) and on how scalable the experiment is across other products (on the horizontal).
Looking at each opportunity through this lens helps his team prioritize initiatives by identifying where they can get more “bang for their buck,” so to speak.
Step 2: Identify the levers and channels
Once you’ve identified the opportunity, the next step is to reverse-engineer a list of the levers and channels that’ll allow you to influence and manipulate that part of the funnel.
For example, if your goal is to increase trial registrations, there are multiple levers you can pull based on where the biggest opportunity lies. You can test the trial signup page, deploy nurtures to bounced users or use ads for retargeting to get them back to the page.
In the following example, we’re examining a funnel from Page Loaded to First Login. I decided that the biggest opportunity lies in testing a couple different variations of the registration form.
Levers to be pulled can be identified by looking at the data specifically for the registration page. Generally, levers will always tie back to a user behavior metric, time on page, bounce rate, fields completed, etc.
For example, I believe that reducing the time on the registration page will lead to more trial signups. Okay, how do we test that? A couple of options are: get rid of non-essential fields, multi-page registration flow, etc.
Step 3: Identify the stakeholders
As you begin to get a clearer picture of the exact goal you’re trying to accomplish and which needles you need to move to get there, you’ll be able to identify the relevant business units, partners and stakeholders.
For example, if you want to run a simple A/B test on a landing page, you need to get buy-in from marketing and the page owner, and potentially some other folks. Once you’ve identified these people, start thinking about how you’ll get them to help with your initiative.
Step 4: Do your research
At this stage, you have all the information you need to create a list of questions you need answered before you can move forward. You should tap the relevant subject matter experts for any pertinent qualitative data, and then—based on that data and all the other information at your disposal—you can craft some qualitative assumptions about the outcome of the experiment.
Step 5: Collaborate and align
Now you’re ready to host a discovery meeting with the stakeholders and partners. The goal of this meeting is to collaborate and align. You present your growth experiment hypothesis along with your high-level quantitative analysis and qualitative assumptions. Following discussion, get the team to agree on a measure of success that’ll guide your actions going forward.
A note about choosing your “North Star” metric: While many say an experiment’s primary KPI should be specific to the product in question, at the enterprise level you actually want to pick a KPI that’s applicable to multiple products.
Following these steps is necessary in an enterprise organization because they help you navigate all the gatekeepers and other obstacles that might otherwise sabotage your efforts. It aims to create partnerships and collaboration to facilitate a more streamlined and efficient execution. And it gives your growth lead the opportunity to keep strengthening key relationships across the company.
Technology stack: Less about tools, more about access
Today, two things drive growth efforts: the people who develop the strategies, and the tools they use to execute them.
What’s in your tech stack depends on the scale of your organization and the types of growth strategies you’re after.
For IBM—a large company pursuing many different kinds of sophisticated experiments—the tech stack includes platforms and solutions to handle a wide range of tasks, including:
- Pipeline Data Governance (CDP)
- Customer Relationship Management & Sales Opportunity Data
- Product Intelligence & Visualization (including heat maps and session recordings)
- A/B Testing
- Customer Success
- Customer Feedback
Each company’s technology stack will be unique, but there’s a universal growth technology challenge that all organizations face. Because there isn’t one team that owns the permissions for all these different tools, the functionality within the overall tech stack becomes very siloed.
This isn’t just a technical issue in terms of integration, it’s a relationship issue. Silos create a sense of ownership. And when you have ownership, people naturally become territorial—not a great thing for collaborative partnerships.
Navigating these kinds of conversations is one of the growth lead’s key roles. There’s a need to be sensitive, tactful and respectful. The growth team can’t just go charging in and demanding access—you have to make sure everyone has a seat at the table and feels like they’re part of the same team.
Technology is a necessary and powerful component of a successful growth experiment, but it only works when the people managing it work well together.
Trust: The real bottom line
Working well together is the foundation of enterprise growth efforts. When you boil it all down, the growth team can’t accomplish much without earning the trust of the rest of the org. This requires strong leadership, bulletproof process frameworks and the right technology.
One way to gain the trust of partners and peers is ensuring you have good data. Good data is critical because we all rely on the numbers to prove our hypotheses and show what’s working and what’s not. If your data becomes skewed, people will start to distrust your analysis, and your credibility will go out the window. After that, it’s all uphill—people will question everything you say.
Zoltan is a passionate advocate for thinking through your data foundations—what you track, how you collect it, your data governance and so forth. You need a strong, flexible, scalable data architecture that can grow with your company and accommodate the fact that people track data in all different places.
As a growth lead, you must be the single source of truth, meaning that you need a solid handle on all the data—and that data has to be accurate. That’s key to keeping everyone on the same page and moving confidently forward in the same direction.
And when it comes to growth, having a united coalition that’s working together toward a common cause is really the bottom line. Managing growth in an enterprise organization isn’t all that different from managing it in a startup, it’s just that you have more of everything to manage—more players, more stakeholders, more technology and more options.
Success depends not on having wildly different strategies or tactics, but on having the ability to manage those strategies and tactics and get everyone on board, aligned and believing that you’re going to succeed.
And you know what? You will succeed.
Improving processes, adding more documentation and holding a bunch of training sessions won’t scale a product-led engine—it’s org design that matters.