The risk of RFPs: Finding a better way to purchase in the AI era

Everyone involved in enterprise software buying knows the (prolonged and often painful) drill: Send out requests for proposal (RFPs). Wait for responses. Sit through dubious demos. Undergo months of internal signoff. 

Today, the risks of RFPs are more than a lengthy timeline and tedious steps. RFPs pose a real threat to your business, especially when you’re considering a customer data solution to help you break through the noise and drive revenue as soon as possible.

Thanks to artificial intelligence, software is now changing so fast that the plodding pace of RFPs can tangibly slow down enterprise innovation, limit competitive advantage, and stunt AI-driven revenue growth. RFPs no longer fit the bill. 

Buyers need to move and make software decisions more quickly — and they need a new purchase process that’s as innovative as the tools they’re considering buying. The way forward? Pilots and proofs of concept (POCs).

Why the RFP process can no longer keep up

Most enterprises stand by their traditional RFP workflow because it’s how they’ve always bought software. The RFP process can take six months or more, plus another six to 12 months for implementation. In all, 12-18 months often pass from the time buyers decide they need a solution to when they actually see value. 

When a software offering (like a CDP or CRM) is fairly static over time, extended RFP timelines can be a slog, but they’re not necessarily detrimental to the business. That’s not the case anymore. 

In the case of AI, though, technology is moving rapidly. Take OpenAI as an example: In December 2023, a request for OpenAI’s gpt-4o to build a prototype website returned a response of “I’m unable to directly create websites.” By June 2025 (just 18 months later), giving the same prompt to model o3 generated a React prototype and interactive preview. Today, modern foundation models can often produce shippable production code that only requires standard review. The pace of acceleration is increasing.

AI industry leaders are setting an incredibly high bar for the pace of innovation, and countless other tech companies are following their example, including the enterprise software vendors you want to buy from.

Technology is changing so quickly that buyers who stick with a traditional RFP process will be a cycle or two behind the industry innovation by the time they get up and running with a vendor. Meanwhile, their competitors who decide to pilot a solution can test the technology and decide quickly — when all is said and done, they’ll be 16 months ahead of peers who trudged through RFPs.

This new world demands a whole new way to source, evaluate, and purchase software solutions.

How a pilot or proof of concept differs from RFPs

The RFP process is designed to support hypothetical decision-making. The enterprise buying committee gathers as much information as possible on the existing vendors and solutions and puts each of those options through internal filters. No wonder RFPs are so time-consuming. 

A pilot or proof of concept flips the RFP model on its head. Buyers can get up and running quickly: Usually, within weeks users can touch and feel the software on their own data or synthetic data that’s structured like their own. Armed with clear deliverables and goals, they’re ready to test and confirm that the software can do what they need it to do. 

Rather than moving forward based on hypothetical results, buyers can see verified value through a pilot or POC and gain greater confidence about choosing a particular solution. 

Even if the buyer is bound to the RFP process, pairing that information-gathering with a proof of concept means collecting newer information (and proven results) that can seal the deal with execs and decision-makers. The result is a shorter decision timeframe and a proven path to ROI.

How to safeguard data during a pilot or POC

Is your leadership hesitant about pilots or POCs with a CDP because they don’t want to test a not-yet-vetted solution with production data? After all, the RFP process puts solutions through their paces with legal to ensure the vendor can even be trusted with that data. 

Ask vendors if you can provide them with a schema of your data so they can generate synthetic data that looks and feels like yours. During the pilot, you can bring your goals and KPIs to the proof of concept—with your same column names and layouts—and test in an environment that feels like home.

4 (more) reasons to pursue a pilot or POC

Faster and more trusted software decision-making gives pilots an immediate edge over RFPs. A pilot or POC wins for these additional reasons, too:

Skip the demo tricks 

Every buyer has experienced it: The hour-long demo that fakes impressive pre-baked results or fast-forwards through a process that takes hours in production. The hands-on testing of a pilot sees through all of those loopholes and gives you ground truth on the product's performance. 

If you’re testing a product that lets you build audiences, a demo could easily fake a real-time result. But in a proof of concept, you get the real product experience of testing on your own.

Verify results at scale 

You often can’t tell how much data is actually being handled during a demo. And to pressure-test enterprise software like composable CDPs, you need to be sure it can handle your company’s volume of data. 

That’s why providing your data schema and specs and asking the vendor to generate synthetic data offers invaluable intel. You can confirm that the product handles data at the scale you need with reasonable response times and performance.

Answer the right questions 

Buyers don’t want to know that an enterprise solution can answer someone else’s business challenges. They need to know that it can answer their questions and solve their problems. 

A pilot lets you bring your challenges and questions to the interface and watch it handle them — questions like, “Can this tool help us create an effective loyalty program and campaign for our highest value customers?” or “Can it help us effectively retarget churned audiences?” With synthetic data populating your structure, you can quickly see how the product or model interprets these questions, the steps it takes to answer, and decide how confident you are in its ability to work for your production data.

Confirm deployment speed firsthand 

A proof of concept that uses your schemas can give you confidence in quick deployment (or not). If a pilot or POC takes a month or two to go live, that’s hypothetically the fastest you would be able to go live after signing a contract. Implementation will only get longer from the pilot rollout, so you now have a baseline expectation for the shortest possible timeline to get your team actively using the product and seeing value.

If you want to avoid software buyer’s remorse, a pilot or POC is your direct line to faster and better decision-making in the age of AI. Stay in step with your competitors (or one step ahead) by moving quickly; blink, and you’ll be left behind.

2 key ingredients for a successful pilot

A pilot gives you greater detail and depth of product knowledge than an RFP ever could on its own, so your definition of success can be more detailed, too. I recommend doing two things before you start a pilot with potential vendors:

1. Pick an extreme pain you have today — one that you expect this software to solve for you. Write down a single-sentence description of what it would look or feel like for you to solve that pain. 

2. Identify a massive business opportunity that you aren’t tapping into today. Describe how you believe the right solution would help you reach that goal.

This two-step process gives you specific use cases to test during the pilot and creates a clear definition of success. If your testing during the pilot helps you make measurable progress toward solving that problem and reaching that goal, you have proof that the solution can help you succeed. 

Evaluating a composable CDP pilot in practice

Let’s say that you’re a large retailer with a retail media network. Your business sells customer audiences to its suppliers. The problem? It takes weeks to get those audiences to live in the system where brands actually buy access to your customers — that’s painful for your retailer and its supplier customers. 

This current state creates missed opportunities, too: Every day that brands ask for audiences and don’t get them is a day they could (and do) buy those audiences elsewhere, even if your data is the best option.

To solve this problem and make the most of that opportunity, you undergo a pilot with GrowthLoop. Coming in, you know the baseline time that it takes you to get audiences live in your system. The GrowthLoop team generates synthetic data based on your schemas, and you can use that data to test and measure the difference in the time it takes to get audiences into the system with the new solution.

If you noticeably decrease that cycle time, you resolve pain for your suppliers and make more revenue quickly. By the end of the pilot, you have verified proof of potential ROI and greater confidence to make a decision — leaps and bounds beyond what an RFP could give you.

Next steps: How to help your org move from RFP to POC

Making the switch to a pilot approach might seem out of reach, especially when you need to convince your execs to embrace something new. “RFPs are just what we do” may be their motto, but it doesn’t have to stay that way. 

Take these three steps to start shifting your organization toward a pilot or POC approach that can keep up with the current pace of innovation:

1. Understand the RFP. Before you can convince your leadership to switch to a new way of purchasing, make sure you understand why the RFP is happening in the first place and which gates are fundamentally necessary. Who mandated it? Why does the process look the way it does? This foundational knowledge goes a long way because it points you to the right person or people to talk to within your organization about embracing pilots instead.

2. Work with partners to tell the right story. Pick a vendor or set of vendors you can work with on a proof of concept that can get up and running and prove value quickly (even if it’s alongside an RFP). Use the pilot to relay the opportunity cost of waiting until the RFP process finishes. Recruit your partners to help your decision-makers see that the POC was a success and that others in your space also believe in that vendor.

3. Find your internal champions. Start by convincing your direct leader of the value of a proof of concept with a goal of working your way up the reporting line within your company. You need to get the right story up to the right executives; your leader (and their leader) can help you do that. Execs want to drive revenue now, cut costs, and deploy AI in a meaningful way. An efficient, effective POC just might be the proof they need to see that a certain vendor reaches those goals — and that a quicker decision (and quicker time to value) is more important than longstanding processes like an RFP.

The case for moving away from RFPs is a strong one. Keep the opportunity cost front and center. Show your leadership what they risk with lengthy purchase processes that lag far behind AI innovation cycles. And remind them what they can gain through faster, more adaptable approaches like pilots: revenue and business growth at the speed of AI.

The new normal for enterprise software purchases

AI has fostered virtually unheard-of rates of software innovation, and buyers need to make informed decisions faster. But should the new pace of play steer buyers away from long-term contracts as well? Not at all — as long as the system is open in how it handles your data. 

Long-term contracts with a closed system (like a traditional CDP or marketing suite) tie you down. If you don’t like their AI tools or roadmap, you’re out of luck for the years ahead.

What about when you choose a composable solution? After you’ve mapped your data to the business objects in their system, you’re now linked to their AI roadmap, but you maintain control over your data. You’re primed to continue innovating with that vendor as their tech evolves, and you have the freedom to bring in new tools, models, or partners that help you reach your business objectives. 

Open access to how the data changes over time gives you freedom for AI innovation because you can build the additional workloads you need.

As you look for vendors to partner with long-term, choose solutions that come to your data cloud vendor; these composable platforms offer the flexibility and control you need to continue innovating. Invest in tech that uses rapidly evolving AI capabilities to shore up your weaknesses and double down on the strengths that set you apart from competitors. These are the vendors you want on your team for the long haul.

The GrowthLoop Compound Marketing Engine is designed explicitly for the AI-speed world — and we offer pilot and proof of concepts to help you test for yourself and make the right vendor decision sooner. From there, our composable solution helps you innovate in the gaps. It enables rapid iterations to drive compounding growth and writes back to your own data cloud. By choosing partners and platforms that are architected for agility, leaders can compound growth instead of fighting fires. Connect with our team to get started with a pilot.