Eliminating Stovepipe BI and Empowering Business Users
by Ron Powell
Originally published November 14, 2016
Matthew, can you provide a brief overview of Colony American Finance?
Matthew March: Colony American Finance was founded in 2014, specializing in investor financing for investors that are interested in buying non-owner occupied houses to get into the rental market or, as is very popular today, to fix and flip.
In our space, access to capital is challenging for a lot of our customers. Banks don’t traditionally lend for investor-owned properties above a certain number of houses. So we provide financing from a bridge credit line, single-asset as well as term loans for investors that previously only had access to capital from expensive source, such as hard-money lenders. We just cleared $2 billion in funding this month so we’re pretty excited about that.
One of the things that would get everyone excited is that you got to start with a blank piece of paper to solve your business intelligence needs. Tell me – what was that like?
Matthew March: I’ve been in IT since the mid-eighties and have been a C-level executive for the last ten years at several different companies. I joined Colony American Finance about a year ago. As I said earlier, the company was founded in 2013 but really hadn’t made a lot of strides from a technology perspective. It was a greenfield opportunity. For the first time in my IT career, I was handed a blank piece of paper. I got to architect and build it, and the only thing they asked was that I not screw it up.
We looked at some of the applications they already had. They were already in the cloud. They were an Office 365 environment. Email was through Office 365. They had Salesforce and some marketing applications, but they did not have a business intelligence (BI) solution.
Since the early 2000s with the information explosion, I’ve been really focused on the BI space from a technology perspective. If you look at the title chief information officer, the most important word in that title is information. Most companies today don’t build anything anymore. They don’t manufacture any products. They generate a tremendous amount of data, and aside from their employees and customers, their data is their most important asset. When they are able to take that data and turn it into actionable information that can impact the bottom line or the top line, that’s where the magic happens.
So at Colony American Finance when I came aboard a year ago, it was very typical. Different company but same challenges. Folks were acquiring data manually out of systems, working with Excel, for example, and trying to de-dupe data and match/merge and bring all the data together to do analysis for marketing, do analysis for sales – to put together very basic, rudimentary operational reports rolling up to the executive team. So time to market to get information was very labor intensive. It would take a long time. For example, the marketing department had several different data sources, and because they had to log into each of those data sources, pull the data down into Excel, and do these lookups, it took them over a week to put together marketing reports for very simple campaign management – looking at lead tracking, ad spend and so on.
So, looking at that holistically and the fact that they were 100% cloud-based put a unique twist on the RFI/RFP for a business intelligence platform. I wanted to stay in the cloud. I didn’t want a data center. For the first time in my IT career, I didn’t have to worry about five-nines on my data center. I didn’t have to worry about outages with my data center. I didn’t have to worry about disaster recovery and geographically diverse data centers, or hot recovery, or cold recovery or replicating data. It turned into a vendor management play. So whether I looked at Microsoft with Office365 or Salesforce or even Domo from a business intelligence perspective, were we selecting the right vendors with the best-of-breed technology, do they have SLAs in place, are they following ISO 27001 27002 standards for information security, do they have SOC 1, SOC 2, SSAE 16 annual attack and penetration testing, and DR testing? This puts the responsibility back onto the vendors we do business with, but it is our responsibility from a governance perspective to make sure we select the right vendors and we manage them accordingly.
When you looked at the needs of your organization, obviously cloud was very important. The ability to support a wide user base – was that a key concern? And time to market?
Matthew March: I think if we look at business intelligence today, it has really changed quite a bit from the traditional data warehousing approaches, design and development platforms of the past. Those platforms are still out there, and at previous companies I worked with SAP, Cognos and MicroStrategy. From a data warehouse perspective, with the Ralph Kimball federated data model design with star schema, we spent a lot of time up front just trying to acquire, cleanse, extract, transform and load the data into a business-centric data model. At Colony American Finance, we didn’t have the time to try to create a business-centric data model from 27 different data sources.
We’re a smaller company. I went from a 76-person IT team at my last company to an IT team of 3. My challenge was how could I empower the business with access to their data without having the bandwidth, the time or the money to deliver all of the 27 different data sources as quickly as possible.
So being a BI veteran, looking at the cloud, did you have concerns about having your information out there now that you’re using Domo?
Matthew March: Not anymore. I think when the cloud first came out, as is typical of all new technology, you want to be on the cutting edge perhaps but not on the bleeding edge especially for information security and data protection. For example, I just came from the banking environment that is highly regulated, and many financial institutions utilize the cloud. Also in the medical world with HIPAA and other regulations, a lot of that data is on a cloud now. So when we looked at moving our data and information to the cloud, again it came back to vendor management. How are they protecting the data? What are they doing for information security? What information security standards have they adopted? What perimeter defense and depth do they have in place? What are the information security tools that I have as a customer at my disposal to do things such as encrypt data at rest and do secure FTP. Those are all important questions to ask during the RFI/RFP process.
We broke down RFI process into four categories:
When we’re looking at a vendor, it is those four components that we need to address to make sure that we’re making the best decision.
I’ve seen BI implementations that can sometimes be very difficult for end users to work with the BI system. How are you finding ease of use with Domo?
Matthew March: Well I always joke that the world runs on Excel, and every company that you work with has Excel in the organization. Excel is sort of the ubiquitous tool, and it’s the technical capability of a lot of the power users within the organization that are pulling together, providing and disseminating the information back to the management team and the front lines.
So when we look at a tool, it is very important from a business intelligence perspective that we select a toolset that not only provides the functionality the business needs, but also allows the business to use the tool themselves.
One of the holy grails of business intelligence is teaching users to fish for themselves. Having worked on many different platform initiatives, Domo is the first example where we’ve actually been able to empower the users to adopt the tool to leverage the data in a very safe, secure environment and to do ad hoc analysis themselves – to create dashboards, cards and reports that have business value without having to rely on IT or the business intelligence developers to do so. So we extended the capability and the capacity and the resource availability to create reports and dashboards by leveraging the Domo platform. It is very easy to use and very intuitive. I think it’s a testament to Josh and Chris who founded Domo – the user experience for any business intelligence platform or any application of technology is paramount to adoption and success.
Can you give an example of where the end users took initiative and solved a problem without putting a burden on IT?
Matthew March: There are two use cases that I’ll share. We actually have investor decks that we update KPIs and metrics on a monthly or quarterly basis. We have a team of analysts that are responsible for doing so. Previously they would download that data from a variety of systems and sources. They would manipulate the data in Excel, make pretty graphs and charts, and then they would cut and paste them as images into our PowerPoint presentation. But that lifecycle is very tedious. And then the same KPIs, the same data, month over month, quarter over quarter.
With the advent of Domo and providing access of all of our data sources with the Domo platform, one of our analysts had an epiphany. He said, “This is an amazing tool. Why are we not using this to automate all of these graphics on KPIs for all of our investor presentations? So every time that we need to refresh them, instead of having to download all of the data and go through a week of work, I could just log into Domo and export it right into PowerPoint.” So in lieu of coming to IT to ask them to develop these decks and dashboards for us, we actually worked with the analyst team, showed them how to use the tool, showed them how to use the data, and they went out and created all of those components for the investor decks. So it used to take a week for a team of 3, and now it’s automated and at any point on any given day, you can log in and export that data into the investor decks.
Another compelling example is where you empower the business with the data that they can actually take action on and leverage the information through the Domo platform to determine correlation, causation, and take action without any IT involvement.
A good use case example was an email campaign that we were working on to get approved to send out. Normally at Colony American Finance, we send out our email campaigns at the beginning of the week. We ended up not getting approval until mid-day Friday to send out a campaign. Our president logged in to Domo on Saturday morning and found that the open rate for the email campaign was horrible. Looking historically at the performance from the dashboards in Domo, the open rate was ten percent of what it should be. In lieu of waiting until Monday to call IT and ask them to have the BI team pull some data or call the marketing team to have them pull some data, she used the collaboration feature called Domo Buzz. It’s like sending an email, but you’re right there within Domo. She notified our VP of production that it just didn’t look right. Our VP of production logged in, leveraged the data in Domo, did some analysis, and came back with the correlation/causation that historically our email campaigns perform better if we send them out at the beginning of the week. The thought process behind it and the aha moment was that our customers generally don’t work on the weekends. They’re 9-5, Monday through Friday kind of people. They’re not in the office. On a weekend, they’re not looking at any emails we send out. When they come into work on Monday, that email is buried so far down in their email inbox that they wouldn’t even see it. So Monday rolled around, and, sure enough, the open rates still looked terrible. We decided to resend the email campaign on Tuesday. Same email campaign. Same customers. Now it would be in the top of their inbox. Lo and behold, the open rate went back to where it should be. The click-through rate went right back to where it should be. And the cool thing was that in our old methodology before Domo, it would have been a month before we knew that email campaign hadn’t initially performed well. It would have taken us that long to compile all the data. And by that time, it would have been too late to take any action. We’d be so busy compiling reports for the next iteration that the data wouldn’t be actionable.
This is a great use case, and it was all without any IT involvement. I didn’t get a phone call. I didn’t have to work on the weekend. We provided them with a tool and the data so they could fish for themselves, see an anomaly, and answer the question without any IT involvement. It’s spectacular.
Matthew, you had an extraordinary opportunity starting with a clean slate, but there are so many companies out there that already have business intelligence infrastructures. They struggle with getting the information to their executives. What would be your advice to them? Is the Domo platform an option for them?
Matthew March: Domo is an incredible platform, and you are right – I was handed a once-in-a-lifetime opportunity. The companies I have worked with in the past – as CIO at the Banc of California and as CIO/CTO at Carrington Mortgage Holding – had existing infrastructures. Every company has the same challenges. When coming into an environment where you’re inheriting an existing architecture – maybe you have an existing legacy data warehouse and existing reporting tools – I guarantee if you did a customer satisfaction survey you’d find the business is not getting the value they need or expect out of that platform.
So how do you transform from that environment to a Domo-type solution where you can empower the business? If I look back at my previous company, we leveraged the Microsoft BI platform to deliver a business intelligence solution. It was a very traditional implementation – star-schema data model running on SQL Server, leveraging SSRS and Power BI. But time to market was very, very slow. And because it took us so long to deliver the data to the business and once we did deliver the data, part of that development tool was then dedicated just to keep the lights on as our source data changed. We had to work with our other vendors to make sure that when they did releases, it would not impact our ETL processes, etc. So what happened was that in other areas of the business where we were not serving their data needs and not providing the solution they needed, they went out and got their own solutions. So then suddenly we were competing with Tableau or Excel. We were competing with all these different reporting solutions.
So now part of the BI team at that company was spending time defending the accuracy of the numbers produced out of the data warehouse. Also, they had to spend time explaining why the reports coming out of Tableau from the real estate division sales team were not accurate. That’s stovepipe BI.
But if I look back at the opportunity with the advent of a tool like Domo, you can introduce Domo to an organization that already has a legacy BI platform and might have another tool. The power is that even though you might have two platforms from a technology perspective that need to be supported as you transition over, you can bring all of that other data into Domo much faster, much more quickly, in a safe secure environment and eliminate stovepipe BI across the organization. You can then leverage Domo to be your data warehouse. You can leverage Domo to be your ETL tool. But you can also leverage Domo to be your presentation layer.
So if you have a legacy data warehouse that you spent a tremendous amount of time developing that has value for the organization, but you don’t have the time, bandwidth, resources or time to market to deliver all of the other data sources that your business needs, I would challenge you to bring in Domo as a presentation layer because you’ll be able to bring in all of your other data sources very quickly and empower the business by leveraging Domo as the presentation layer for the new data sources as well as the legacy business intelligence infrastructure on a go-forward basis.
There is a way to bridge that chasm. There is a way to transition that. And with the licensing model that Domo offers, it’s very, very cost-effective. When I look back at the Banc of California, we never brought in any of our marketing data. The BI team never had the bandwidth to bring in all of the marketing data that was needed to make decisions or to do campaign management and see the ROI on the various marketing campaigns. They were focused on the originations platform. They had it on the banking platform. That’s all they could do. That entire team of four people was 100% dedicated to doing that.
So by bringing in a tool like Domo, we could easily bring in all the other data sources and empower other business divisions with the tools and the data that they needed in a safe and secure environment – and without a huge maintenance footprint.
At Colony American Finance, we rolled out 27 data sources in two months. And we delivered in two months what typically took over three years for me to deliver historically. And I did it with one developer. Currently, one developer spends three percent of his time on maintenance items, and 97% of his time either teaching the business how to fish for themselves and cross-training more users to actually leverage the toolset to create their own cards to do their own ad hoc analysis or now we’ve moved into the predictive analytics space and leveraging a third-party application called Big Squid to provide some predictive capability leveraging our data.
Another area that is really key is data quality as well as process and data flow of information across an organization. How did you address that?
Matthew March: Data quality is key for any business intelligence initiative, and aside from the tool selection, one of the keys is to go out and do data profiling across all of your data sources – garbage in/garbage out. And when we talk about data quality, everybody thinks it is an IT issue – a technology issue. But at the end of the day, the business owns the data. Data quality issues are also owned by the business. But you want to leverage technology to build a better tomorrow. So if we look at leveraging Domo as a toolset to do the statistical analysis across all of your data sources to look at the data quality, you identify where there are areas of opportunity upstream to correct those issues at the source.
Traditionally that’s very challenging in business intelligence because the BI team is so far removed from the application development team. And if they submit requests to the app/dev team to update the legacy application to enforce data type, precision and so on, it will never happen. But from a cloud-based environment, we’re able to identify in Salesforce, for instance, where we have configuration opportunities to make fields required to have drop-down lists of values, to enforce the data type precision up front. So fields that are needed downstream by our capital markets team that are required are being populated as part of the normal workflow.
We can also leverage Domo as a platform to look for operational efficiencies between different systems to identify the bottlenecks. With Domo, we can leverage business intelligence from an operational effectiveness standpoint, maybe from a Six Sigma DMAIC pipeline perspective, to look for areas of opportunity to improve the processes. Or maybe there’s a manual step that can be automated. By looking at timelines and pipeline management within our business intelligence, they can be measured. That’s important because if you can’t measure it, you can’t manage it.
One last question. We’re hearing a lot about predictive analytics and machine learning. Does that apply to your situation as well?
Matthew March: It absolutely applies. It is the first time in my career where we’ve been successful in empowering the business to fish for themselves. I’ve worked on many multi-million dollar business intelligence platforms, and we were never able to achieve that goal. And we’ve done this. We’ve had Domo now for ten months. And we’ve been working with business lines to automate their reports, we have brought in all of our data sources, and I have been teaching people to fish for themselves. So now we’re able to free up resources to take it to the next level.
You hear a lot about machine learning and predictive analytics, and while we leverage business intelligence tools from a human perspective and use the data visualization tools, you can see that there are trends and anomalies, but we don’t always understand why. We might not always be asking the right questions. So from a use case perspective and a pilot that we did with Domo and Big Squid, we were able to better understand our customers and their behaviors. If someone applies for a loan, what is the probability of them actually funding with us? Or even better yet, if I have a list of 30,000 investors, if I knew what the key attributes are, can I target those investors that are likely to do business with us? I can really customize and make some effective campaigns to go after all 30,000 of them, but for that segment of that population of opportunity where there’s a high probability of them actually doing business with us, you bet I’m going to customize a marketing campaign to go after them in a different way than I would go after that entire population of 30,000.
Great. Thank you Matthew for enlightening us about Colony American Finance and what you’re doing with BI and how you’re serving the decision makers within your organization.
About Colony American Finance
Colony American Finance (CAF) is a specialty finance company that provides a range of debt products to residential real estate investors. The company offers portfolio and single asset term loans for stabilized rental properties as well as short term credit lines for acquisitions. CAF was founded in 2014 to finance single family, townhome, condo and small multifamily properties for customers nationwide. Its products are tailor-made for investors and it provides attractive rates, rapid timelines and closing certainty. The company works directly with borrowers as well as with brokers and correspondent partners. For more information, visit www.colonyamericanfinance.com.
Recent articles by Ron Powell
Copyright 2004 — 2020. Powell Media, LLC. All rights reserved.
BeyeNETWORK™ is a trademark of Powell Media, LLC