Serif Health

So today we’re chatting about Serif and how it’s used at your company. Before we jump into that, could you give a brief overview of the company and your role there?

We’re an enablement platform for primary care practices. Our goal is to unite these independent practices into a network, taking care of their back-office services and supporting their transition toward value-based care and improved contracts with insurers. We’re a small, early-stage company. I’m the COO, but I also currently oversee operations, customer service, and product/technology.

What was the original need that drove your search for a product like Serif?

It was twofold. First, as we begin negotiations with insurance companies, we need to have a clear understanding of what other providers in the market are being paid. This information is crucial when asking for a higher reimbursement rate. Second, having market comps serves as a powerful marketing tool for approaching practices. It allows us to demonstrate to them that they aren’t being paid as well as their competitors, so it’s sales enablement data.

What were the key requirements you looked at as you were evaluating products in this space?

We primarily focused on acquiring commercial payer data, as most of our patients are covered by commercial insurance, and the majority of insurers are in that space. Our main requirements were good coverage of commercial payers, clean and easily interpretable data, and above all, accuracy and reliability. We wanted to be able to compare the data against our internal data points to ensure it was of high quality and suitable for benchmarking purposes.

Unlike most people who prioritize optimization for large health systems, we were interested in the “long tail” of practices. These smaller, independent practices at the bottom of the stack needed our attention, and we wanted to understand how they are paid and how similar practices are compensated. And that became our decision point as we were looking at the two main vendors. One made it easy to access all the data for every practice, while the other didn’t provide an easy way to access detailed information about the long tail of practices. The second vendor provided aggregations that omitted the specific details we were interested in. Although the data existed, they would have required us to request a custom and expensive solution to access it.

Who were the vendors that you considered, and how did they compare to each other?

The two vendors we’re familiar with in this area are Turquoise and Serif, though there may be others. We initially considered going it alone, since the payer data is somewhat public but not easily usable for non-experts. However, we quickly realized that managing it ourselves would be overwhelming and require significant financial investment in data analysts and engineers. And after seeing what companies like Turquoise and Serif could give us, we realized why companies like this exist. They’re able to take publicly available data and present it in a more digestible way. So we quickly abandoned the idea of doing it ourselves.

If you’re looking for quick aggregations and information on payment rates, such as a particular insurer in a particular state for specific procedures, Turquoise has invested more in the user interface to allow for easier searching. Serif, on the other hand, may be less developed in terms of the UI, but they were able to work closely with us to understand our specific needs and deliver customized information. Turquoise was less flexible and couldn’t fulfill our request for a validation exercise. They were also not as clear about whether they had the long tail data on small providers that we required. Additionally, Serif was more cost effective, probably two or three times cheaper than Turquoise.

How did you find the sales process overall with Serif?

Serif was amazing. They had a small team, and the founder was involved in the sales process. We had a couple of meetings where they took the time to really understand our needs. We were able to give them specific details about what we wanted, and they came up with a customized pricing plan that fit us perfectly. There was no pressure to buy more than we needed or fit into a pre-packaged deal. We got exactly what we wanted at a price we were comfortable with, which we really appreciated. And they delivered the data to us in just two days. The team at Serif was great to work with.

What was the onboarding and setup like?

We didn’t bother using the main Serif UI because it didn’t prove to be useful, and I don’t think many other customers use it either. However, Serif delivered the data files to us in a user-friendly format, allowing us to easily load them into our own data warehouse database. This made it convenient for us to manipulate the data as we desired. It wasn’t like a typical EHR onboarding process.

Can you describe your workflow with Serif?

We wanted raw data files, and we ended up with a massive amount of data, well beyond what a spreadsheet could handle. We had a data engineer spend about an hour transferring the dozen or so files into a database. This allows our team to access the data through a BI tool, which allows sorting, filtering, aggregation, and other actions. We now have a stable system in place where we can query specific things, like lists of providers in a certain zip code. We’re considering developing internal tools or even using this data as a marketing tool, like creating a lookup on our website for people to compare rates provided by different providers in their area. However, we’re not at that stage yet.

Currently, we use the data to answer questions in meetings with payers or practices. We want to understand the median rate that payers are offering to other providers in a specific market, compared to what they’re offering to our providers. This information helps us determine how to present our rates in pitch decks and highlight any discrepancies with what competitors are receiving.

We received the data two months ago, which should be sufficient for the next six months since we’re not yet in contracts. However, once we start securing contracts, we’ll need regular updates on market trends. Currently, a static data set is acceptable. However, we anticipate moving to quarterly or even monthly data refreshes in the future.

What do you see as areas of growth for Serif?

With both data providers, they enrich the data they receive from payers, but mostly it’s just regurgitation. It would be helpful if they could do more translations. For example, a payer may provide five different rates for one practice, with everything else being the same except the rate. We’re not sure how to handle that. We’re at the mercy of the payers regarding the information they provide.

Serif gave us a few tips to navigate this. Usually, if we encounter multiple rates, we take the highest, as it’s often the most accurate. However, we still have to figure things out on our own. What I really want to know is the true source of what this provider currently gets paid by this plan when they see a patient. It would be great if they could guide us better to find that answer.

Another challenge we face is that, since their data comes from payers, the information only includes the practice’s NPI number and sometimes their name, or the name is enriched from another source. It would be really great if we could understand the affiliations of different practices or providers. Currently, we have to create our own rules for identifying affiliations. If they could create a taxonomy, such as labeling an independent practice associated with a health system or an IPA, it would simplify things for us. We might be one of the few customers who need such specific data, but that’s where we struggle right now – enriching the raw data with information that helps us find those answers.

I would like them to assist me, as a customer, in connecting more dots. Currently, they mostly clean up and regurgitate the payer files. It would be excellent if they could incorporate other data sources that make the information more meaningful to save work on my end.

What would you characterize as Serif’s key strengths?

It’s highly customizable. Really great, responsive team. Very usable data. Well structured. Easily delivered. And I think they have very broad data availability. I think they cover a huge amount of the market.

Is there anything that Serif offers that you are not currently using?

We don’t use the main Serif UI because it doesn’t surface the data that we need. Instead, we prefer to access the raw feed directly. The Serif UI seems to be designed for simpler use cases, which doesn’t align with our specific needs. This seems to be a common issue with many data providers – they create a product that works for the majority but doesn’t always fit everyone’s requirements. Considering that we end up filtering the data and using a CSV provided by them anyway, I don’t think it’s worth investing too much time in the UI.

And they offer regular updates and the option to send the data into a data warehouse or wherever we’d like, but for now, it’s not something we need.

How do you find the account management and support?

I think the support team is the founding team at this point, and I have to say they’ve been amazing. There are countless payers out there, some of which cater to very specific markets. Although Serif might have data on these niche payers, they didn’t include it in the initial data dump because of the sheer volume it would add. However, if we come across a medical practice that deals with a significant number of patients covered by a particular Medicaid MCO, I simply reach out to the support team and ask if they have data from that provider. In most cases they do, even those in niche markets, and they’ve managed to retrieve those files and deliver them to me on the same day. They’ve been incredibly responsive and helpful and have met all our needs without fail.

Looking back, do you think you made the right assessment in going with Serif?

Yeah, totally. We’re super happy. They’ve been great.

Do you have advice for other buyers who are evaluating vendors like Serif or Turquoise?

Based on our experience, if you have specific data needs that go beyond hospital data, you’ll likely require a vendor that can provide customized solutions or conduct targeted research. Most of the available products are designed for institutional users, so choose a vendor that is willing to accommodate your requirements. It shouldn’t take them much time or effort; they just need to be a little creative. We found it extremely beneficial to work with someone who thoroughly understood our needs. Also, be prepared to invest a lot of effort on your end to achieve the desired outcome. In general, even when dealing with these vendors, having access to a data engineer is essential.

Do you have any advice related to actually using the data to achieve better contract rates?

Confidence is key. We’ve found it helpful to verify the data Serif provides against our internal data to ensure its accuracy and reliability. We also use certain methods and approaches, such as taking the highest rate and removing outliers, to refine the data. It’s important to invest some time in validation before presenting the information back to a payer, as it may have been supplied by them or on their behalf. This way, we can confidently stand behind the data we present.