Pressing the reset button: An agency’s digital transformation journey to bridge the data gap

We had been chasing a “single source of truth” before our realization. One system to give us everything we needed. This was what we were after from 2016 to 2018. We were aiming to shoehorn everything into our AMS to give us a single source of truth. This was wrong.

3 months ago   •   9 min read

By The ReFocus Team

NOTE: The views expressed here are those of the author(s) and do not necessarily represent or reflect the views of ReFocus AI.

The article that follows is based on the lessons, insights, and ongoing activities from an agency that has pressed the “reset” button a few times with systems and strategies. For color, we are a 58-year-old firm which when this all began in 2016 had been underutilizing our Agency Management System (“AMS”). Until then, we only used our AMS to its full potential in four areas: premium receipting, certificate producing, renewal report running, and policy file note taking.

So, we pressed the reset button. “Phase 1” of the digital transformation reset took 36 months and included a complete accounting delete and reconstruction. Phase 2 is still in-progress.

Through all of this, we developed three “scopes” to look through when changing processes, setups, or practices:

  1. How does this affect the business team’s ability to carry out their sales and service?
  2. How does this affect accounting?
  3. How does this affect reporting?

And for #3, we realized something big in 2019. We successfully tested and validated our realization in 2020 and currently, in 2021, are in the middle of designing what will be our long-term solution.

AMS serves a significant purpose–it’s the engine of the agency’s systems. But you need more than an engine to make a vehicle.

That realization? Do not rely on AMS for your full business reporting solution...

We had been chasing a “single source of truth” before our realization. One system to give us everything we needed. This was what we were after from 2016 to 2018. We were aiming to shoehorn everything into our AMS to give us a single source of truth. This was wrong. AMS serves a significant purpose–it’s the engine of the agency’s systems. But you need more than an engine to make a vehicle. Similarly, you need more than an AMS to get the most out of your data.

The solution is a separate data facility and reporting tool - platforms that are specifically built for these purposes. Use an AMS for its front-line agency servicing strengths. That is where it excels. But use a data platform for its data strengths.

The outcome of this project will enhance our freedoms for #’s 1 and 2 while enabling a whole new set of possibilities for #3.

The Problem(s)

AMS data is really good, raw data. “Good” in terms of it represents the “on-the-field” activity. And “raw” in the sense that it needs a lot of reclassifying and re-categorizing. There are many situations in which the best way to process in AMS causes some report integrity issues. For instance, policy and transaction downloads from carriers are crucial, but situations arise like moving a policy from one insurer to another that creates reporting challenges. The expiring carrier often downloads a cancellation. The renewal carrier downloads a new business transaction. From the carriers’ perspectives, these explain what happened. But from the agency’s perspective, this is not a one-in, one-out scenario. It’s a remarket (or requote, or rewrite, whatever your agency likes to call it). The downloads do not talk to each other to reclassify the cancellation-new business combination to a remarket.

In fact, this issue was the primary driver of a problem we had last year. Our Lost Business and New Business reports for 2020 produce numbers which compared to our actual financial results for 2020 were off by a factor of 2.  2x wrong!  And this is after a three-year accounting reconstruction. Makes it hard to want to run a Lost Business / New Business analysis.

Another issue: scoping and the prerequisites to most AMS reports. Take production reports. These typically depend on production information, which generally involves accounting statements like direct bill entry statements being entered and closed. We have our own perspectives on the direct bill statement entry process (to be addressed another time) but in short, this causes the report to lag and be based on what external parties (insurers) tell the agency, not what the agency already has in its own system.

You can produce reports out of AMS; it will give you numbers on a page, no doubt. But if you drill down and try to understand what is actually happening, you run into some pretty sizable constraints pretty quickly.

In summary, it’s actually quite difficult to understand what is happening at the agency, account, and policy levels through off-the-shelf AMS reporting, particularly in real-time. You can produce reports out of AMS; it will give you numbers on a page, no doubt. But if you drill down and try to understand what is actually happening, you run into some pretty sizable constraints pretty quickly.

The Solution

We are building our own databases in Microsoft’s Azure environment. And we are mapping out Power BI dashboards that will summarize the data as we want to see it. Here is what this enables:

  1. More control: report on what we want to see, not what the AMS limits us to seeing with off-the-shelf reporting
  2. Ability to apply “treatment” to the activity in our underlying systems (correcting for the scenarios like the remarket issue mentioned above)
  3. Consolidate information from multiple business systems
  4. Better [actual] forecasting

Control

We were long ago inspired to design our reporting around the commission drivers for agencies. Those being 1) new business, renewal, and cancellation transaction sums and counts and 2) changes to rates and exposure bases (insureds’ property values, payrolls, revenues, etc).

And maybe this is where our passion for solving all the above started, but in our view, reporting on these factors out of our AMS was inadequate. Not only did there seem to be quite a lot of transaction correcting necessary, but the access to the rating and exposure information is minimal.

The important thing here is that by building our own data facility, we can control how we handle the data and how we report on it. While the rate and exposure measurements are difficult, ultimately it is information that the agency possesses. And what our data facility absolutely enables is the ability to capture, store, measure, and report on any of the data that we already possess. And it allows us to do this without the constraints that exist with trying to report on these things “off-the-shelf” in an AMS. We have the data. We design how we capture, ingest, and analyze. It’s not the easiest thing in the world, but it’s simple.

Treatment

Not only can we correct the straightforward issues like the rewrite scenario above, but we can actually create our own transaction types based on what they measure (e.g. cancellation(insurer X) + new business(insurer Y) = remarket). We can add more granularity to the transactions. For example, we could break lost business down by cause. Maybe there is a premium increase? Maybe there is a material increase in the cancellation rate due to failures to comply with inspection recommendations? Maybe there is a reason to break out cancellations as we correlate it with claims denials?

We can correct download bugs by applying sequencing rules and logic to delete duplicates, apply mathematical operations, or reclassify where necessary.

We can compare commission percentages paid against commission percentages expected via commission schedules.

Ultimately, this alternative method of reporting allows us to find, create, and tell stories now!

And we can combine analyses. We can break out remarketed situations (which have already been reclassified based on cancellation > new business sequence logic) where we 1) saved the client premium dollars 2) while increasing our commission.

Ultimately, this alternative method of reporting allows us to find, create, and tell stories now!

Consolidation

Merging the information from the various sub-systems may be one of the most critical features of a data facility. There are two significant improvements through this capability:

  1. Enable deeper access to the data points that add real insights
  2. Provide the freedom for the sales and service teams to do their job better

As mentioned, reporting on rate and exposure elements is challenging if relying on our AMS alone. This is despite the agency having access to this information. While many AMSs have data fields for this information, accessing these data points via any sort of meaningful report is limited. By connecting AMS data to quote data by consolidating information from not only the AMS but also from our underlying quoting systems (which can be multiple), we will have a much better handle on not only how changes to insured values and exposures are happening, but how those are affecting and/or are likely to affect our revenues.

Regarding enhancements to how our teams work, creating a data facility that acts as our source of truth allows us to forgo selecting systems solely because they work well together (unless it’s truly what we need from a functional perspective) and actually choose systems that give our teams what they need. This is particularly powerful when considering all the quoting systems it seems to take for an agency to quote all of their markets these days.  While it remains ideal to avoid having too many quoting systems, with a centralized data facility agencies can afford to adopt what they need to deliver the best products to their clients because the activity and underlying data is captured, stored, and accessible via the facility.  With that being said, a key criterion for procurement becomes whether the systems allow an acceptable level of access to the information contained within. If it does, we can extract that information into our data facility to both improve the quality of our data and better enable our business teams to do their job well!

What other information can be consolidated into the facility?  CRM data, email activity data, phone log data. We can consolidate everything and where possible, cross reference to accounts and policies to open a new realm of servicing level measurement. Lifetime customer value has historically been difficult for agencies to measure.  Not so much anymore!

The other dataset that is a part of many new analyses enabled by the data facility…commission schedules. And while the premium drivers above are still valid, it is really important to go deeper and track commission schedules too (especially these days). Similar to the calculations of rates against exposures driving premiums, premiums against commission rates drive commission dollars!  We do not stop with assuming premium changes will drive commission changes - we factor changing commission schedules.

Forecasting

Forecasting is totally possible based on the information we already have access to. We read a statement recently that agencies were using their end-of-year production analysis to forecast the following year’s cash flow. While this can make sense given existing systems and processes, obviously much can change over the course of a year. There are likely some trends missing from your forecasts if you are aggregating cash flow assumptions this way! For instance, new business vs renewal commissions from carriers. Particularly if your agency grew heavily in new business this year. It is not uncommon for commission schedules to drop renewal commissions by 3-5 points. If your agency grew significantly by producing a lot of 15% new business commission rates with carriers that pay 10% next year, there is a 33% change to commission rates on the horizon.

To us, these are things we felt we should be able to see and budget for, but could not in the conventional reporting environment.

And in terms of AMS reporting, we never felt like we could understand what revenue to expect over the next three months. In fact, we gave up on a budgeting project in our AMS because it felt silly going crazy budgeting expenses when ultimately we really didn't know what to expect with commission revenue receivables. And each time we tried to use the prior year’s revenue, the timing of direct bill payments, changes to commission schedules, or unfactored new business, policy cancellations or non-renewals would lead to a material variance. To us, these are things we felt we should be able to see and budget for, but could not in the conventional reporting environment.

With our data facility, we can populate a report to show what we expect to receive in commission based on last week’s new business, renewal, and cancellation transaction activity. It’s not perfect, but it’s better than any other method we ever tested. And it at least allows us to check direct bill statements against something on our end. We hope we can actually develop this part of the project into a full scale receivable management system, so our balance sheet can benefit from this too!

Final Thoughts

To conclude, our advice is to stop relying on your AMS to give you what you want to see and/or stop expecting it to be the single source of truth. It is extremely valuable as an AMS. You need it for that. But build a report overlay to:

  1. get more insights from your AMS,
  2. add and cross reference your other system information,
  3. report on things however you want to report on them!

Special thank you to the author of the article, Judson Norton. He is the Principal at Erwin Insurance Agency, an advisor to multiple insurtech startups, and a passionate space technology proponent. If you would like to learn more, he can be reached at j@erwinins.com or through LinkedIn.

Spread the word

Keep reading