The power base within enterprise organizations—and with it, the way vendors sell into these enterprises–is shifting. Much of this is attributable to the consumerization of IT: Workers now bring the expectations of user-friendly consumer tech to the tools they use at work. They won’t put up with clunky, tough-to-use products when they know that user-friendly, intuitive alternatives exist.

This trend has unlocked new budgets inside organizations for solutions tailored more for individual employees, as opposed to large groups or company-wide rollouts. Now, employees are no longer limited to using rudimentary, legacy tools like Excel spreadsheets; many can opt for new cloud-based and/or open-source tools to help them work more productively and effectively.

This shift has prompted many startups selling enterprise software to pursue what’s called a product-led-growth (PLG) strategy. This basically means that if you make a product that’s appealing enough, you can gain adoption at the grassroots level, turning individual employees into evangelists and relying less on formal (and often expensive) sales-and-marketing programs to drive your company’s growth, at least in the initial stages. It seems to make sense, given current tech-buying trends.

But behind the PLG hype is a little-known truth: Even the most successful product-led strategies can hit roadblocks today inside enterprises. Despite the proliferation of easy-to-adopt technology tools and more buying autonomy for some individual employees and small groups, enterprises are still not free-for-alls when it comes to technology sales. In many cases, oversight from powerful IT, security and legal departments—particularly in the current work-from-home environment–is still important, and this can trip up startups who don’t understand these dynamics.

Simply put, enterprise organizations still make decisions based on inputs from multiple groups. Just try to sell a workplace-collaboration product without getting the blessing of the enterprise’s cybersecurity team, for example, or expand your reach inside an account where a PLG competitor has been busy managing up the management chain, in addition to selling across small groups. You won’t get far.

Some technologies still require an all-or-nothing deployment. For example, as the value of big data—and the need to comply with new regulations—grows, many new data solutions, particularly those working as official systems of record, need to be chosen and championed by the C-Suite. There might be less oversight of more-niche data products, such as tools for data visualization, where authorization to the underlying data is handled elsewhere and executives would be fine with employees choosing the best tool for their specific job. Other products, like the ServiceNow platform used by large companies to manage IT services, would also likely be difficult to implement on a piecemeal basis.

Still, a great UI, one hallmark of products sold through a PLG strategy, really does matter at enterprises. Recently I heard a Fortune 500 company CISO compare two solutions, noting that the product from the Battery portfolio company he ultimately selected was chosen for its great UI. In his words, the competitor’s product “looked like a third grader built the interface…why would I pay for that?”

The bottom line is that selling into enterprises looks different depending on how your product is actually adopted by internal groups, and by whom. To succeed, you need to understand the power dynamics informing buying decisions today, which remain very important. That starts with answering two big questions.

Question 1: To whom are you selling?

If you’re selling into a cost center like IT, HR, legal, and so on, you should first recognize that cost centers are where the majority of technology is procured inside an organization. That’s a good thing, by and large. That said, it does mean that annual budgets, shifting priorities and, likely, a legacy process or tech stack may create some switching costs for your target users.

The second challenge: Cost centers don’t expand the company’s top line. Your solution might be great, but you still have to sharpen your pencil and prove why your buyer needs your widget over hundreds of other competing priorities. Even if your technology is an open-source or SaaS solution, that doesn’t necessarily mean an individual contributor can simply swipe a credit card because they like your software. In most organizations, this employee will need to jump through a lot of hoops before they can adopt a new tool. You’ll need to enable security, compliance, collaboration and more to get budget-approved dollars.

This isn’t always true, but it’s true more often than you might suppose. In a recent conversation with Goldman Sachs, we asked how easy it is for a developer there to adopt a new software testing tool, a market defined by PLG companies such as Slack, Atlassian’s Bitbucket, HashiCorp’s Vault or Selenium. Without missing a beat, the Goldman executives replied that systems are actually heavily locked down – this despite everyone believing in the current bottoms-up market for software tools.

Selling into a cost center in a top-down manner means you are enabling a spreadsheet jockey to be leaner, more compliant, more efficient and so on. Founders should focus on landing the head of the department they’re targeting, but your sales department will need to engage the rest of the team to fully close the sale, so everyone’s on board with the purchase.

If you’re product-led, you’ll want to frame your value proposition as ‘we make your life easier’ in some way. It’s relatively easy to identify the right buyer persona, but to land the sale, you’ll need to arm those buyers to answer questions they’ll encounter from legal, procurement, management and so on.

If you’re selling into a product group, the buy decision likely will be based on a cost-benefit analysis. The phrase “I could build that in a weekend” has been uttered more than once. The group will need to know what it would cost them to build a solution from scratch, not just in terms of actual dollars, but in terms of time, speed to market and ongoing maintenance– and then compare that to the cost of simply buying your tool. Your job is to help them do this math.

When selling to a product group, your target buyer is the product manager or engineering leader. To win over this buyer, you’ll need to ensure you’re aligned with their product schedule. You’ll also need to remove technical barriers to adoption, such as APIs, excessive documentation and so on, to avoid an engineers’ rebellion. You’ll also need to build a financial model that scales with the customer.

Question 2: Are you ready to sell to the enterprise?

Whether you’re selling to a product group or a cost center, enterprises will have certain expectations based on the stage of your company.

If you’re a seed stage or Series A company, an enterprise is betting on your domain expertise and agile team of technical all-stars. Ultimately, they’re betting on the founders. In exchange, they expect support—and will want you to prioritize building out features they need.

One strategy is to sell to early adopters. You can build a nice business selling to other venture-backed businesses, who don’t have legacy tech stacks and are nimbler. You can also sell into the mid-market at this stage. You should track two key metrics at this stage: time to deployment / usage and how many new customer logos you’re adding. A mid-size enterprise or digital-first organization is likely a better fit than a bigger enterprise, but the leadership team (i.e., not commissioned salespeople) should spend enough time in sales conversations with big, regulated titans to understand when you’ve got your product that’s acceptable to them. Whoever your customer is, at this stage you should expect to spend significant effort building credibility, even if your founders have a long track record.

If you’re a Series B or C company, the risk in continuing to focus on venture-backed startups is that your mid- to late-stage company hasn’t yet built its enterprise muscles. Make sure you have a list of recognizable enterprise customers who have cleared security, procurement and other hurdles. You should start to think about adding some product features that you’d consider nice-to-haves, so you’re building beyond the bare-bones essentials. Corporates will expect to find a relatively full-featured product from a company at this stage—and a team capable of managing enterprise sales and offering the support they expect from major IT vendors.

Whatever stage your company is in, remember that all enterprises are political, and you have to touch multiple constituencies to sell.

Navigating the enterprise organization can be tricky for startups. Your culture is so different from your customer’s culture, you can run into a lot of unexpected hurdles if you don’t understand how your buyers’ minds work and what political issues they’re dealing with. But if you approach an enterprise in the right way, you may land that big account that could help take your company to the next level.