It started several weeks ago. I missed a call with the panel of people preparing for a FundRaising Success webinar. Well, I wasn't there to stick up for myself, and we all know what that means. I got the topic that no one wanted — yep, I got assigned the topic of modeling relative to how the nonprofit industry is pushing toward integrated, multichannel marketing.
Of course, my topic was only supposed to last 10 minutes and it went perfectly, but in the meantime I just had this nagging feeling that I needed to dig deeper. The more I thought about it, the more I wanted to know what the experts were thinking in this area — so I decided to just ask! For the most part, what is in this article are direct comments/quotes from the folks that I believe are the brightest and the best in our industry around this topic.
I asked four questions of each of these people.
- Richard Becker, president, Target Analytics
- Jeff Birkner, director of marketing intelligence and strategy, Eleventy Marketing Group
- Caity Craver, CEO, DonorTrends
- Jim Emlet , principal and co-owner, Integral
- John Ernst, chief strategy officer, nonprofit, Paradysz
- Brent Eskew, executive vice president, Wiland Direct
- Yannis Kotziagkiaouridis, vice president, market analytics, Merkle
- Gretchen Littlefield, president, InfoGroup Nonprofit
- Chuck Longfield, chief scientist, Blackbaud
- Curt Swindoll, executive vice president, strategy, Pursuant
- Andy Wilder, vice president, analytics consulting group, Epsilon
(Today's post presents their comments on the first two questions. Check back tomorrow for the rest!)
On some topics there was very similar thinking, on others —not so much. But hands down, the perspective from every one of these individuals/agencies is worth the read as everyone in our industry tries to understand how to be better marketers and fundraisers.
1. When thinking through all the challenges the nonprofit industry is facing these days, where do you think modeling can be the most helpful?
Brent Eskew: Cooperative databases are a huge benefit when it comes to solving the prospecting problems we have. Using the traditional application of modeling to optimize net files following the merge/purge process just prior to mailing is a great strategy, and the names that fall out can be replaced with the next highest scoring cooperative database names.
Curt Swindoll: The state of our ability to capture data relevant to the relationship development process has never been higher. We need to think in non-financial terms, particularly around behavioral data that speaks to interest and engagement.
Richard Becker/Chuck Longfield: Nonprofits have been accumulating significant donor data in their CRM systems for years. However, the vast majority of it goes unused when making key donor decisions. The world of donor interactions/engagements is avoided, as are their preferences, relationships, demographics and interests. To take advantage of more data, practitioners need to use modeling. So, since all of this other data is valuable, there isn't an area where modeling couldn't help. Models would select better populations, target better messages, set better ask amounts, and identify more optimal channels. The lack of modeling is holding back the state of fundraising significantly.
Andy Wilder: The creation of a single, enterprise-wide view of each constituent through the consolidation of data and information from many disparate sources has offered fundraisers a new perspective on their constituents. The challenge now is to leverage all aspects of a constituent's engagement with the brand to maximize the value of the relationship over time, rather than the profitability of a single campaign or channel. Transforming awareness into action requires a rigorous and empirical understanding of the potential value of each constituent. This knowledge, made possible through predictive modeling, lays the foundation for successful CRM and enables the organization to determine the appropriate level of investment and the optimum level of communication for each donor.
Caity Craver: The power of the model comes from understanding how to apply the intelligence of the score. The nonprofit industry needs to catch up — and fast. Modeling is used by over 90 percent of the industries worldwide to protect the bottom line. Fundraisers are still stuck using the same "human brain assumptions" to select and segment donors.
Yannis Kotziagkiaouridis: Digital media and channels have increased the complexity of how constituents are consuming information and making decisions. To be successful, nonprofits need to understand the impact of their fundraising activities and use that information to allocate their budgets in creating relevant constituent experiences. Modeling plays a key role on every single step of this process including:
- An understanding of the complex relationships across media and channels
- A better distribution of marketing resources.
- An understanding of audiences and the creation and delivery of relevant marketing experiences to the most valuable segments
John Ernst: Modeling can be most helpful in donor/revenue growth and donor engagement. Growth has been elusive for many nonprofits, especially those heavily reliant on direct mail. Gone are the days of one offer to one audience and trying to optimize for that response. Modeling will be critical to helping identify, define and develop the audiences (note plural!) that have the best revenue growth potential and which ones don't. Even more important, they'll be instrumental in quantifying that potential, because it's going to take investment to pursue them. Donor engagement has been affected by the fact that channels are integrating at rapid speed for the direct-response market as a whole, and the nonprofit industry isn't immune. The classic RFM-based modeling principles work great when you control the promotion history/schedule and work in a single channel. But try to calibrate for a mail piece landing in home to a donor who may have visited the website via their own accord last week, saw a DRTV ad, and received an e-mail event invite from the local chapter — it gets a bit more complex! Modeling will be required to leverage the latest data feeds and map out what are the best combinations of channel, timing and message to communicate with your donors.
Gretchen Littlefield: Modeling has already had a huge impact on new donor acquisition, but its potential has not been fully realized yet. For example, for quite some time now, smart marketers have been modeling within the merge process to eliminate the prospects least likely to respond. Now, the really smart marketers are modeling to reach the prospects most likely to make a second gift. It's not enough to bring in new donors in droves; the donors must give more than once. Simply stated, modeling offers the ability to maximize budget by impacting both sides of the equation: improving performance and decreasing cost. Its greatest impact is in reducing promotion cost while simultaneously improving either net or gross revenue — or both.
Jim Emlet: While modeling can be brought to bear on most marketing-related challenges, the two areas where it is perhaps most helpful are new donor acquisition and donor retention.
2. Do you think modeling is underutilized by the industry and if so, why do you think that's the case?
Eskew: It is definitely underutilized in acquisition yet has advanced further in retention strategies. Why? Historically modeling in acquisition was unnecessary — vertical lists and exchanges produced acceptable results. Additionally, as the economic model for new donor acquisition started to change, some opportunists moved onto the scene with expensive pay-up-front and trust-me models that produced hit-and-miss results. Interest and demographic data —age, income, and "interest in skiing" and "outdoor activity" for instance — often is simply not predictive enough information to pinpoint new donors willing to begin giving regularly to an environmental cause. Modeled names work better. Both acquisition and house file names selected and ranked from best to worst based on an individual's "spending behavior" (transactional history) and other sophisticated indexes found to be meaningful for each participant achieves better results. Mailers can select the segments they wish to mail, or eliminate unresponsive mailing segments, thus reducing mail cost and increasing overall mailing performance, which neither traditional list acquisition selection nor RFM house file selection can accomplish as effectively.
Swindoll: It is most definitely underutilized. Why? First, it's easy to continue practices that have worked in the past, even when those practices fail to produce the results they once did. We stop asking questions about how we might reinvent our fundraising approach and processes, including new ways of engaging donors, new ways of measuring results, and ultimately new ways of capturing and evaluating information related to our most valuable donor relationships. Modeling is used all the time in successful commercial businesses. But those same successes are rare in the nonprofit world. Nonprofits find what works and stick with it.
Becker/Longfield: Unfortunately, vendors have not provided such great tools for this. CRM systems don't allow you to write your own model nor easily plug in someone else's model. I'm not sure you want nonprofits to write their own models, but even if the industry started with the plug-and-play option it would be helpful. But, in the end, nonprofits need to get better at tracking the non-transactional data in their CRM system, and vendors/agencies need to get better at identifying the non-RFM data in client's systems.
Wilder: I think the nonprofit community will always march a few steps behind the commercial world when it comes to innovative strategies. However, in the past decade the industry as a whole has picked up the pace. Predictive modeling, which was once viewed as a novel accessory, has evolved into a requirement in all RFPs that cross my desk these days. Nonprofits have long understood the benefits of collecting and acting upon the information stored in their databases. However, a number of organizational, structural, and philosophical barriers have prevented many fundraisers from experiencing the full potential of sophisticated modeling solutions. With the recent trend of executives from other commercial verticals assuming leadership roles in the nonprofit sector, I think we'll continue to see the industry close the gap.
Craver: The industry is separated into different populations. The established nonprofits have been mis-using and over-utilizing models for some years now. The newer smaller organizations have not been using models at all, and the mid-size nonprofits are somewhere in between. Mostly the reason models are not affectively utilized is because there has been a gap between the science and the application of it. This is primarily because most nonprofits do not have mathematicians or statisticians on their staff. If they do, rarely are the mathematicians or statisticians brought into the strategy discussions. Additionally, where modeling is being applied, it's often misused without clear direction on the objective of the model or the cost benefit to applying the scores. DonorTrends estimates less than 4 percent of nonprofit organizations use scoring systems to target their donors. Until there is more widespread adoption, fundraisers will continue to select donors using their "human brain assumptions."
Kotziagkiaouridis: Nonprofits have been successful using models for targeting communications and getting a better understanding of their constituents. The opportunity we see today is for nonprofits to utilize advanced statistical techniques to better understand the impact and interaction of all their marketing activities, offline and digital, and attribute to them a proportion of the credit for the final desired outcome. The result is a better distribution of marketing spend. Unless organizations better understand multichannel interactions, there is little they can do to optimize them. Merkle believes the inability to connect offline and digital behaviors at the constituent level and the constraints of internal data and technology integration are the major factors that prevent organizations from maximizing the potential of modeling.
Ernst: In a traditional (direct-mail) sense, I think modeling is actually being overused by the industry. Most data companies, agencies, co-ops, and even Paradysz — we're all trying to optimize response for our clients in this tough economic environment. And, we do a pretty good job of it, which helps our respective clients' business. But, it is actually hurting the overall industry. Since most organizations are pursuing donors using similar lists, and using similar techniques (annual fund packages, labels, notepads, etc.) the collective effect on the industry has been an "over-targeting" of the same audiences, resulting in lower response rates and shrinking list universes for everyone. What seems like an obvious and noble effort for an organization has ultimately resulted in a real, negative effect for the industry as a whole since we're not bringing new names into the marketplace or identifying new, productive audiences. Modeling should be used to find these new, non-traditional or undeveloped audiences.
While we're still being challenged to optimize our clients' programs, we're also seeing much more of an appetite to grow and a willingness to invest in new techniques to pursue those goals. Our next-gen models are looking for audiences that look different than the traditional ones. It's not just about predicting response, it's about informing that certain groups should receive a different message, ask level, or be spoken to in a new channel. It's also about identifying audiences that don't show up on traditional direct-mail lists and using that model to leverage new data sources to find them. It's about finding that some audiences are much more valuable and require greater upfront investment while others need to be divested.
Littlefield: There's a misconception that modeling will "cost more money." In reality, modeling makes marketing more efficient, relevant and cuts costs. In the commercial sector, co-ops and prospect databases are commonplace. Why aren't more nonprofits taking advantage of prospect databases or co-ops? One reason for this under-utilization may be due to cost concerns, or it could be due to the fact that many boards of directors and marketers see themselves as stewards of their organizations' donors and there is a general reluctance to share data. As a result, the nonprofit sector continues to trail its commercial counterparts and potentially miss the opportunity to spring board their acquisition programs. It's a complete paradigm shift.
Emlet: Very large organizations have seemingly embraced the practice of using models to inform their direct-marketing decision-making. However, the vast majority of organizations certainly underutilize modeling. In general, modeling is perceived as complicated, expensive, and requiring data that is not readily available. For the most part, these concerns are unwarranted.
Vice President, Strategy & Development
Eleventy Marketing Group
Angie is ridiculously passionate about EVERYTHING she’s involved in — including the future and success of our nonprofit industry.
Angie is a senior exec with 25 years of experience in direct and relationship marketing. She is a C-suite consultant with experience over the years at both nonprofits and agencies. She currently leads strategy and development for marketing intelligence agency Eleventy Marketing Group. Previously she has worked at the innovative startup DonorVoice and as general manager of Merkle’s Nonprofit Group, as well as serving as that firm’s CRM officer charged with driving change within the industry. She also spent more 14 years leading the marketing, fundraising and CRM areas for two nationwide charities, The Arthritis Foundation and the American Cancer Society. Angie is a thought leader in the industry and is frequent speaker at events, and author of articles and whitepapers on the nonprofit industry. She also has received recognition for innovation and influence over the years.