When it comes to holding an association event, there are a lot of things you don’t want to leave to chance. From your keynote speakers to the flow of traffic, predictive analytics can give your organization a better experience by making sure that your members are getting what they want from every session of your event.
Just because things seem to be going back to a business-as-usual status doesn’t mean that associations have stopped having events. Even though there has been a necessitated shift toward online events rather than in-person, this is a huge part of many associations’ marketing and membership strategy. More than 97% of marketing professionals believe that hybrid events are here to stay, with another 71% planning to continue maintaining a virtual audience once they return to physical events. And even online-only association events, like SURGE from Sidecar, saw huge numbers in 2020—over 1600 attendees.
The goal with either style—online or in-person events—is to create a positive experience that brings value to your members. And the best way to do this is by using your data effectively to identify the motivations, needs, and interests of your membership.
Because so many associations have been forced into virtual or hybrid events, there is a lot of data that has been collected—maybe more data than you know what to do with. Consider all the places where your membership has engaged with you over the past year. This might include:
But what exactly can you do with this data to make your event better? There are actually several areas where good data can create a better event experience for both you and your members.
How can you make sure that you are inspiring the right people to come to your event unless you know what motivates them? Through good data analysis, you can begin to break down your membership into cohorts that will help you better target you marketing to both new and existing members.
RELATED ARTICLE>>Member Success: Association Members are Happier with Tailored Member Journeys
Take a look at your historical information from past conferences. Which were the most attended? Which speakers had the best reviews? Which sessions had the most social media engagement (both on your social accounts and across the social media sphere)? This information will give you a good idea of the high-quality content you should offer this year.
It’s a simple thing to look at how many people attended and shoot for that same number, but using your data wisely means looking at what didn’t happen and solving the problems of those that didn’t attend. For example, if you see a huge number of people that attended in 2019 but not 2020, or noticed that people who hadn't purchased anything from you in 3 months are much less likely to register, this allows you to create personally targeted campaigns to focus on those who are likely to register or may be more on the fence than others.
As you move back to in-person or hybrid events, historical data from previous in-person event years can help you structurally organize your showroom area and other key meeting spaces. For example, if your data highlights vendor demand for a certain area of your exhibition hall, you can consider making that priority placing and price accordingly. If some yearly speakers saw huge numbers of participants, it’s vital to get ahead of that traffic concern now and place them in spaces that are not only large enough, but provide members access to membership renewal or purchase opportunities.
Even though things are slowly allowing us to connect with members face-to-face, there are a number of considerations that need to be made when preparing for events in the coming year. Your data, when leveraged correctly, can give you the leg up to provide the best event experience—both online and offline—that lead to better engagement and membership retention.
If you’re interested in how Tasio works with association leaders to use data to solve problems, or are interested to find out what kind of data you need to solve problems like these, check out our webinar “Is My Data Good Enough?”
You can also contact us directly here.
On January 27th, Thomas Altman and Dray McFarlane, co-founders of Tasio, were featured on the Mary Byers Successful Associations Today podcast. Their topic was “How AI Can Increase Membership,” and included some great information about how associations can use the same technology that Netflix and Google use to better understand their membership and predict behavior.
“We don’t actually have to create anything new,” said Dray, speaking of how Tasio gathers data and predicts member behavior. “We’re surrounded by this technology. When you get suggested movies on Netflix or targeted ads on Facebook. We’re just taking that [technology] and predicting who will renew or not by applying those exact tools to associations.”
One of the things that Mary spent some time on was the fact that Thomas and Dray have used their data science knowledge and association background to bring the power of AI-driven data analysis to a group that has never really explored it before. Specifically, the team explained how associations can be leveraging this technology for applications like:
A key takeaway from the discussion was that, while associations benefit from Tasio’s brand of predictive analytics and member retention strategies, the members themselves benefit as well. With better, more individualized data, associations are primed to offer better services, have more targeted marketing, and make changes to increase the value to their members.
“This is a really powerful tool, I believe, for associations going forward,” Mary said, as the discussion wrapped up. “Thank you for making this so approachable. I’m not in a membership department, but you’re making me wish I was.”
Hear the entire Successful Associations podcast here. Or, learn more about what Tasio does by contacting us.
Part 3 of a 6-part blog series.
In order to eliminate the risks and losses that come with organizational challenges, you first have to spot them coming. While this may sound like white magic, it’s possible today with the power of predictive risk analytics.
Predictive analytics is revolutionizing traditional risk assessment by giving us a data-driven structure for better and more accurate fortune-telling capabilities. By assessing and transforming risk drivers, themes, and behaviors into future insights, we can stop threats dead in their tracks and flip potential catastrophes on their head.
Traditional risk assessment methods for associations are often complex, subjective, and difficult to automate. These traditional (and oftentimes manual) approaches to assessing risk include surveys, workshops, and monitoring industry trends.
But newer predictive analytics approaches are fully automated and provide your association and its board with organization-wide risk awareness (and associated risk assessment) for strategic planning and preventative purposes.
Ranking risks gives you a weighted risk assessment based on likelihood and impact scales. Once you have this insight into the likelihood of a disaster, you can then move forward with your predictive risk management processes (coupled with some good ole’ fashion human interaction) to prevent it.
In fact, the goal of predictive risk analytics is to assess and estimate with a high degree of certainty the operations that ARE and ARE NOT conducive to the success of your association so you can then build decision-making support systems.
Conducting risk assessment with predictive analytics also allows you to prove at any point in time that you’re meeting the expectations of regulators, examiners, and board of directors, as well as your members, investors, team members, and communities.
You should also link risks to activities to prioritize what activities need to be monitored. This prevents emerging threats from harming overall organizational metrics and KPIs, and creates an organization-wide culture around risk-aware and risk-conscious decision making. And culture eats strategy for breakfast.
Membership is the foundation of your association. You can use big risk analytics to automically gauge waning membership numbers and respond by bolstering member experience and targeting.
Using predictive analytics in operational risk management is useful for identifying and addressing threats related to data governance, compliance, efficiency, operational losses, and capital impacts.
You can use risk analytics to empower your organization and its reputation, align strategy and culture, and keep tabs on member and stakeholder confidence. For example, predictive risk intelligence allows you to scan social media and behavioral patterns. You can then use this arsenal of data to map social and industry changes as well as the reasons behind a potential or growing problem to take charge of shifting scenarios.
Using predictive analytics in financial risk management ties in with forecasting. For example, you can use risk analytics to identify and address limited resources by prioritizing resource-building campaigns and initiatives.
You can even use predictive risk analysis and monitoring to detect a natural or manmade hazard that may have negative humanitarian consequences.
Using predictive algorithms, you can determine the likelihood and impact of a hazard in a defined period and the risk it presents. This will allow you to circumvent the hazard by implementing protective structures and other defenses.
Image Source: UNHCR
Here are the different humanitarian-related threat categories:
RELATED ARTICLE>>6 Ways to Use Predictive Analytics to Grow Your Association: Member Targeting
Socio-demographic factors: gender, age, position, education, marital status
Program details: member engagement, interactions, attendance, feedback
Giving behaviors: changes in amounts, seasonal trends, donor feedback
Predictive risk analytics is a process of using data mining, machine learning, and artificial intelligence to make predictions for the future and provide actionable insights on what would happen under given circumstances.
Creating a culture of risk-sensitive decision making using predictive risk intelligence safeguards your association against imminent threats and helps you stay a step ahead of potential challenges.
Want to learn more about how you can assess and minimize risks before they happen? Read The Association Retention Playbook: The 5-Step DIY Membership Retention Strategy and Workbook for more actionable information your association can start using today.
This is part 2 of a 6-part series on predictive analytics for associations. Read the first one here.
As we roll into the new year, planning and strategy is probably something you’re already tired of thinking about. It takes time, energy, and effort to create good financial forecasts. And if you’re creating forecasts based on spreadsheets, spurious or sloppy CRM data, manual sales reports, and your intuition, you might be spending more time and energy than is really necessary.
Forecasting is an invaluable way to plan your financial future and membership growth over any given period of time. But these traditional approaches to forecasting take far too long, cost far too much, and generate far too little insight about potential outcomes.
With predictive analytics, you can have a greater understanding of your association's cash flow, develop more accurate financial forecasts, and spend less time buried in paperwork.
Predictive analytics is revolutionizing the forecasting process by using an objective, data-fueled approach based on predictive algorithms. This type of algorithmic forecasting—also known as predictive analytics forecasting or predictive forecasting—allows you to confidently drive pretty much all decision-making.
So a financial forecast is not a one-time spreadsheet that you dust off and update each year. Instead, it’s a living model of your financial situation. Predictive analytics makes ongoing forecasting possible in the short term, medium term, and long term.
How can predictive analytics help your cash flow forecast in the short term? It allows your association to evaluate upcoming expenditures and revenue to guide tactical planning and programming decisions. Such short-term decisions may include the specifics of member events and marketing adjustments.
In the medium-term, using predictive analytics enables you to take an in-depth look at expenditures and revenue at the end of the year and in one or two years from now. If it looks like there will be a shortfall, you can respond quickly with minor strategic decisions that increase funding opportunities and cut back on spending to avoid dipping into reserves. And if financial projections point to a revenue surplus, you can put more money toward member services and provide added value to your educational offerings.
Now financial modeling in the long term is a whole different animal. A long-term financial forecast gives your association advance notice so you can pivot and implement new strategies. By using predictive analytics for this long-term forecasting, you can pinpoint possible shortfalls or surpluses in three years, five years, and beyond so you can take the appropriate steps to balance your cash-flow beforehand.
If your long-term financial forecasting models predict revenue will be down, you can transition to programs that build cash flow, such as special member-funded projects, to avoid liquidating investments. And if your financial forecast indicates a revenue surplus, you could decide to hire additional employees in specific departments, hold more value-added events and educational programs, and follow through with large-scale investment initiatives.
Related>>Dollars and Sense: What is an Association Member Worth?
Having a clear picture of your association’s cash flow at any point in time depends on the accuracy of your foresights. Here are some internal and external factors to keep in mind to improve the accuracy of your financial forecast.
Another key component to consider is demand data. Past financial performance is a good indicator of future financial performance. So you have to look at historic data and reports. You can then use this demand data to identify which targeting strategies, campaigns, investors, and association employees/leaders contribute the most to revenue.
Accurate financial forecasting is critical for building an enduring association. Developing a financial forecast using predictive analytics can show you exactly where your association is headed, based on past performance and other internal and external factors. From there, you can use that information to prepare for the future and continuously grow your association.
Ready to build a rock-solid financial forecast? Check out our free ebook, The Association Retention Playbook: The 5-Step DIY Membership Retention Strategy and Workbook to discover key insights into how to meet and grow your membership targets and financial position.
This is part 1 of a 6-part series on predictive analytics for associations.
Membership is the heart of your association.
From the kinds of education you offer to the special events you host each year, improving the way people see, use, and interact with your association is key to meeting your membership goals. With member experience so critical to keeping them engaged, it’s likely that the overarching question that’s at the forefront of your mind is this:
“How do we get a better idea of who our members are and how to bring them value from our association?”
Of course, you’ve probably tried to get at the heart of this problem before. You may have developed member profiles based on demographics like age, membership type, or length of activity. Maybe you’ve split them up by how much they financially or otherwise contribute to your organization. Still, what we’ve learned over years of working with purpose-driven leaders is that you probably don’t know how to target them beyond those superficial elements.
Enter predictive analytics.
Predictive analytics offers your association a way to segment and target your members to get a good look at what actually differentiates them—not just what your best guess is. Through better use of your current data, you can build a predictive model that’ll help you see trends in your members; divine (yes, as if by magic) the reasons for their interactions with your association; and then use your new knowledge to build activities, programs, and a personalized member experience that help them stay connected and engaged in your cause.
We used to do marketing based on member surveys, prior purchase history, then, eventually, member pathways to the website and conversion. We could see who buyers were from who signed up for education courses and came to conferences. But these were all responsive tactics to target members that required expensive manual processing and made segmentation of members difficult.
Predictive analytics presents a proactive AND flexible way to target members. Not only do we use data on those same old metrics to predict potential future outcomes, but we also test and try new metrics to suss out subsections of members that we didn't even know we had. Then we can target them specifically to personalize the member experience and build more effective ways to keep them engaged.
Your members are not all alike—far from it actually. And if you don’t understand your existing and potential members, no matter what your targeting strategy is, you won’t deliver a fruitful member experience. You could even run the risk of inundating users with mixed messages at inconvenient times, which is incredibly harmful for member engagement and retention.
So you must understand what makes your individual members tick and WHY. Specificity gives you the flexibility to cater to their needs and desires. This is why segmenting your audience and delivering focused messages and tailored campaigns is essential to truly engage them.
Here are some demographic and online behavioral data to consider to better understand your members:
Luckily, predictive analytics allows your association to use this individual data to automatically segment members with a high degree of granularity, and then devise a more individualized and impactful member experience.
When exploring how to increase the value of your programs and events, it’s vital to perform deeper analysis on the member experience. If you can identify ways to improve how people view and engage with your association, while spending less money, the result is:
Here are some specific ways in which associations can use predictive analytics to accomplish these three ideals:
Targeting is a no-brainer to create a better member experience today. And with the insight gained through predictive analytics, you can better target your members to unlock the true value of your data and make better use of your marketing dollars.
It’s by using this ‘predictive targeting’ that you can retain more members, increase your impact for good on the world, and calculate the future success of your association before it even happens.
Of course, it’s one thing to recognize the value of using predictive analytics to maximize member experience and another to effectively implement this approach. If you’re ready to bridge this gap and start developing real ways to bring better experiences to your members, check out our free ebook, The Association Retention Playbook: The 5-Step DIY Membership Retention Strategy and Workbook.
The truth is, each member of your association is clearly worth more than the dues they pay. They provide man hours, word-of-mouth marketing, and emotional and creative insight that keeps your association moving forward.
But there is a benefit to learning how much each member contributes to the financial goals of your association over the lifetime of their membership.
Once you have put a dollar amount on the monetary value of your members, you will be set to give them better discounts, customized service, and valuable offers to keep them engaged.
We created a simple spreadsheet to figure this calculation out yourself which you can download here:
NOTE: The above workbook is not to be used in the place of professional accounting or financial advice. The Lifetime Member Value worksheet is designed to give you a rough estimate of your member’s worth over the lifetime of their membership. For a more detailed breakdown, speak to your accountant or financial advisor.
There are several things you need to know before you can do this simple calculation. The most important being that you will working with a historical data set—a set of year-long data where the outcomes (lost or retained) of the members is known.
Work with your team to gather this information, which may include IT data, financial data, and employment data. You will need to know:
Total number of last years’ members with member loss data. You will need to create a historical data set of everyone who was retained or lost from a year-long period before now. The closer to today, the better this data will be.
Total dues and non-dues revenue for historical data set. This will be split into two sections on the workbook, and will need to be included for everyone on the historical data set.
Total cost to service a member over historical data period. This includes salary and cost of membership departments, publication, and marketing and promotional advertising costs. It also includes any monetary or financial benefits members receive. It may also include other elements, depending on your unique association offerings.
Average lifetime of membership. You find this by reviewing your historical data set for members who were lost, or “churned.” For each member that was churned over the period of your historical data set, calculate how many years they were members. Then create an average for the entire group of churned members.
For your historical data set, you need to take the total cost to service a member and divide it by the total number of members (including those who left, or were “churned”). This is the total amount of funds that you spent per each member last year.
After you have created your average lifetime for membership, you will multiply this by the average dues and non-dues of revenue that each member generates each year. In the Association Retention Workbook, the calculation is done for you.
We will also do a similar calculation for the cost to service a member of their lifetime with your association. This is done by multiplying the average lifetime of a member by the yearly service cost per member.
By subtracting the cost of servicing a member from the total revenue over the lifetime of a member, you can see whether you are spending too much on servicing or if you have room to offer discounts and other financial incentives to increase your retention.
The revenue minus your costs is your break even point—this is the highest amount of discount you can give before you are making zero dollars towards your association’s goals.
According to the IMPACTS Value Study, a new member to your association brings in an average yearly income of $114 in revenue, with a renewing member bringing in a yearly income of $189 by year 5 of their membership. The costs for retaining these members is inverse, with new members costing between $20-$25 each year and renewing members costing between $4-$5 each year.IMPACTS VALUE STUDY, ColeenDilen.com (10/14/17)
Despite the calculation above, you might realize that there are some reasons why you would be willing to go over the break even point on the behalf of your members. For many associations, there are intangible benefits from offering some programs, resources or discounts.
These are a valuable part of identifying the value of your members, and can be considered as you are using these calculations to make decisions about where to trim or increase your spending.
Every member represents a huge investment of time, effort, money, and resources on behalf of your organization. And while all members are valuable, retained members bring in the most revenue for the least cost.
We have created a strategy to help you use the calculations above to retain your best members through data-driven predictions.
Retaining association members can feel a little like shooting in the dark, especially if terms like “metrics” and “predictive analytics” seem like a foreign language. For many association leaders, member retention is about in-person connection, surveys, and generalized marketing that is hard to really know how well it’s working.
One of the things that we talk about a lot at Tasio is the need for both person-to-person and data-driven strategies for keeping membership numbers high.
By developing the right metrics and using them to develop member profiles, you can better determine who in your current association membership is getting ready to leave—and target them with specific marketing messages to keep them happier.
The big picture idea is that members take action (or don’t take action) based on how invested they are in your association. They take action through things like:
And they take inaction through things like:
The series of behaviors that members do or don’t take contribute to whether they are going to renew or not, which is the basic idea of predictive analytics.
There are behaviors that members take that can help us identify whether or not they will renew. These behaviors are called metrics.
Developing these metrics is one of the most important things you and your association leadership team can do to help identify who is going to renew.
The most important thing to remember is that developing metrics should rely mostly on your own intuition and experience. You know your members the most, and there are some areas of your association that will be specific to you and no one else.
That being said, here are the key steps for creating metrics that really work to predict member loss.
Step 1: Gather Your Team
It’s always a good idea to do this part with a team that you trust. It should include leaders as well as current members and former members (if possible). You’ll want to choose people who have connections to a lot of people in your association to get a good overview of the kinds of services and roadblocks that they are interacting with on a daily basis.
Step 2: Know Your Data Limitations
In order to make predictive analytics work, you need data points for all your metrics for as many members as you can. That means you need to be able to choose metrics for which you have solid data. This data might include:
You don’t have to have all this data now, but you must make metrics that are concrete enough that you can get that data in the future.
Step 3: Brainstorm Metrics (Recency, Frequency, and Monetary)
The goal for your brainstorm team is going to be developing 5-10 metrics that give a good overview of the behaviors your members do or don’t do before they leave your association. We always suggest that you should have at least a few in three key areas: recency, frequency, and monetary.
Recency. How recently did this behavior occur?
Frequency. How often does this behavior occur?
Monetary. How much is this behavior worth?
Step 4: Create a Historical Sample of Members with KNOWN OUTCOMES
This is a crucial part of the process of using metrics to develop member profiles. In order to make sure that your team has chosen metrics that are predictive, you really need to gather a sample size where the outcome of the member behavior is known. We suggest that this is a year-long period that is pretty recent—ending within a few months of the time you create the data set.
Remember, the more recent the data, the better it will be to help you predict future member behavior.
The most important part of this step is that the outcome of the member is known. You have to be able to see the kinds of things a member did before they left your organization so that you can use that data to accurately predict how your current members will act.
Step 5: Analyze Metrics and Remove Bad Ones
It’s unlikely that every metric you choose is going to be a winner, and that’s pretty normal. Once you have gathered your data, you’ll want to put it into a program that helps you analyze how effective that metric is for predicting member loss. You can do this in Excel yourself, or you can borrow our workbook for predicting member loss below.
If you choose to use ours, all you have to do is add your membership data into the “Data” tab and the Excel spreadsheet does the rest of the work for you. You’ll be able to easily see which metrics are accurately helping you predict member loss and which ones should be thrown out.
If you have any questions on how to use the system or would like more information on the Tasio predictive analytics model for retaining members, download our free ebook here.
Part 2 of a 2-Part Series. Read Part 1 Here.
As with most things in 2020, the presidential election feels like it's been going on for decades. Despite the states slowly counting and recounting ballots (as well as the assumptive new president, Joe Biden, making active steps towards a presidential transition), litigation and accusations reign supreme as we enter the second week of November.
In our first part of this post, we talked about a couple of the big missteps that major news outlets made in contributing to this general feeling of confusion. Small sample sizes, incomplete data, and failure to segment were just a few of these. The biggest ones—failure to communicate uncertainty and the impact of predictions on outcomes—are the ones that blindsided most Americans.
Especially in the month leading up to Super Tuesday, there were many polling experts who were giving big odds in Biden’s favor. FiveThirtyEight had Biden as the winner 89 times out of 100 the week before Super Tuesday—pretty favorable odds with none of the nail-biting uncertainty that was to come.
In retrospect, they were right. But more than a handful of Americans were surprised to find that the race ended up much closer than all of these models predicted. But the problem is that these models presented their predictions as gospel, completely ignoring the level of certainty of these wins by not including them in the analysis at all.
In association terms, this would be like not reporting a specific likelihood that the member will not renew, but rather classifying them into groups like "Persuadables, Lost Causes, Sure Things, and Do Not Disturb.” It's important to communicate the output of the model in a way that makes the new knowledge more actionable and useful for your organization.
This is an important element of developing models, not just for presidential elections, but for associations wanting a clear glimpse into how to proceed to create a more sustainable and financially sound organization.
Another takeaway is that sometimes the model can actually change the outcome.
So for example, there's research showing that when people are confident in an outcome, they are less likely to participate in the process that creates that outcome. In a recent study from Dartmouth and U Penn, people were given a chance to “vote” on an outcome for a nominal fee. Before each voting session, they were shown randomized statistics showing different “nominees” ahead or behind in the polls.
What they found is that, the more likely their desired outcome was, the less likely they were to participate again in the voting game.
There are cases where modeling is a dynamic process, meaning that the output of the model itself changes the impact of the inputs to the model. Predictive models should be robust to this kind of process and have mechanisms for updating impact of the inputs.
So for example, in the case of election modeling, you build a model by gathering data on behavior and preferences from voters. Then you create a model that says "Candidate X is 99% likely to win the election." Then those same voters see that model, and change their behavior and preferences because of it. If the model is not built to anticipate that kind of behavior and update, it quickly becomes wrong.
In the case of associations, let's say you build a model that predicts the likelihood a person will buy any product if you send them an email with a discount for the product. Let's then say that your model identifies 30 different products a person is likely to buy if you send them an email with a discount. It might be tempting to then send that person 30 different emails in close succession. But that quickly gets spammy and the person will most likely stop opening the emails altogether.
BUT if you build something in your model to take into account the days since you last sent an email, the model would tell you to only send one email and then wait for the next one.
When creating models that interact with the outside world, it's important to understand the effect of that interaction because it could change the effectiveness of the model itself.
The 2020 Presidential Election has been a rollercoaster of a ride, both from an emotional and predictive analytics viewpoint. As we near the end of it, it is a great time to reflect on the lessons highlighted from the management of the prediction data and make better use of your own data to avoid the catastrophe of a razor-thin, election-sized growth margin.
To learn more about how we help associations with predictive analytics and data management for better member retention and overall strategy, download our free ebook.
Rolling into the end of October, it felt like predictions that were being forecast about the upcoming (now past-tense) 2020 Presidential election were almost fact.
And then Super Tuesday came. And Super Wednesday. And Super Thursday. . .
Although aggregators like FiveThirtyEight were (in the end) correct in their projections, there were also some big areas that they were blindsided—areas that made the 2020 Presidential election more of a nail-biter than anyone was expecting. This has us thinking about what lessons we can learn on behalf of our association clients, namely, that accurate data is crucial to creating predictions that work.
From where we sit, there are five main problems that analysts ran into that made the data going into the presidential election skewed.
These are the same key issues that we see with associations who are using predictive analytics to develop strategies for their organizations.
As Dan Valenzuela recently quotes in his article “What if Stories Are More Predictive Than Data,” a friend of his stated: “How can I trust a poll using only 700 people to predict how a whole [frickin’] state will vote?”
The fact is, while pollsters did their best to represent the demographics of the states in which they were working, there are a lot of limitations to gathering a complete data sample over the phone or in front of Walmarts. There are very human reasons for some people not to want to participate—not enough time, not enough resources, protection of personal space and rights—and the pandemic made this issue even worse.
In addition, we have to take into consideration the infrequency of presidential elections—just over a dozen in the recent polling era. That’s not many data sets to choose from. That kind of infrequency creates a lot of uncertainty, even if the data was more complete.
A more robust data set is the solution. Large and reliable sets of data that cover not 5% of the population (or in our case, your association memberships), but 100%—or at least 99.9%.
For associations, data like how many CEUs, how many purchases made, and whether or not they are accessing the association online portal are like mini-votes they are making (or not making) for your organization. You need to be collecting and analyzing all of them frequently and completely, or else accurate behavior predictions are virtually impossible.
As with a too-small sample size, incomplete data means that your predictions aren’t based in reality and are basically useless for building strategy upon.
A perfect example of this is the polling averages that were reported leading up to the presidential election. Aggregators like RealClearPolitics (showed Biden up by 8 points the week of the election) and FiveThirtyEight (Biden +7) had no concept that the final race would come down to just a few hundred thousand ballots across seven key states—a race that is still too close for some to call.
One of the big issues that was presented even before election day was the concept of “shadow supporters” of Donald Trump, people who were planning on voting for him, but not willing to tell pollsters (or anyone else, for that matter). Regardless of the social or psychological reason for this withholding of information, this was one of the key areas of failure in the accuracy of the predictive models. How can you possibly predict behavior if you’re not getting a straight answer about how people are feeling?
This is another area where associations sometimes are blindsided by their data. Sure, there are exit forms from CEU courses and annual questionnaires about areas like quality of education, cost of membership, value to the individual member, etc. But how many of the members are consistently providing answers? Is it just the few outspoken supporters (or non-supporters)?
Remember, on a typical bell curve, the people most likely to answer those questionnaires and fill out the suggestion box form are the top and bottom 15%—the people who either love you or hate you. That means that there is a full 70% of “shadow voters” who may never tell you how they feel. They’ll just leave.
By focusing on the right kind of data points, you can actually bypass this issue and look directly at actions. Which actions are people who are engaged with your association participating in? Which actions precede a member’s loss? That way, you’re getting the complete picture.
If you were going to host a pre-COVID party and were in charge of drinks, you would be an idiot to buy six 24-packs of the same thing. Why? Because not everyone you know likes the same thing as everyone else. Hopefully, you would have an idea of what percentage of your guests liked hard lemonade versus a local IPA and who would only drink soda or water.
Although analysts for the 2020 presidential election were able to segment generally, there were huge blind spots that they didn’t account for.
For example, Polly, the AI system in charge of the analytics for Advanced Symbolics’ predictions, was wildly inaccurate in predicting the closeness of the election results, giving Biden a 332-166 lead in the electoral college as well as Florida, Texas, and Ohio (all states that voted red).
One of the key problems the AI had was its inability to dissect subsections of voters who might be galvanized by different reasons. In the case of Polly’s Florida prediction, it lumped all voters from south of the border (including Cuba) as “Hispanic,” failing to recognize the huge Republican Cuban contingency.
Associations have a similar problem when it comes to reviewing data. You might see drop-off levels in the over 65 demographic when it comes to your website use (getting on the member area, signing up for courses, etc.) Although you might assume that those members aren’t very comfortable with the technology, you may be ignoring a vital subsection of 65+ members who are very web-savvy but just are choosing not to deal with your outdated or glitchy software.
Accurate segmentation is the key to developing association marketing strategies that connect with your members, and can only be done with the right kind of data points and analysis.
Of course, these three missteps are just a few of the errors that major polls dealt with in projecting the presidential election outcome. In our next post, we’ll talk about how communication, and feedback loops also played major roles in predicting (or not predicting) the future—and what associations can do differently in their own organizations.
To learn more about how we help associations with predictive analytics and data management for better member retention and overall strategy, download our free ebook.
COVID-19 has made 2020 a really hard year for associations—especially those who are paying attention to their membership data. According to Cafe America, 67.9% of nonprofits have seen a decline in contributions with 97.4% expecting to see further declines over the next 12 months due to the impacts of this worldwide pandemic.
Still, there is a silver lining to the devastating impacts of the coronavirus on association membership.
When I work with associations to create in-depth data reviews to create retention strategies, one of the things I focus on is making sure that we have a year-long period where we can see members who were retained as well as those who were lost.
Although this year I’ve seen a lot of losses when I’m doing these data reviews, I’ve noticed that they are the same kinds of behaviors that members were taking before COVID-19. Although they are leaving faster this year, they are leaving for the exact same reasons.
What does this mean for you?
This is a perfect time to set up a predictive data model for the upcoming year because the members who have left during COVID-19 were probably going to leave anyway.
Even though the numbers seem higher, the behaviors that members are taking before they leave are the same as in previous years—they’re just more clearly defined.
Let’s take a look at what we can learn from an example.
In this case study, this state-wide membership association was unable to provide many of the in-person CEUs that it usually did in prior years. Although it eventually managed to transition these to online courses, the engagement numbers were still much lower than in 2019.
Here, we can see that the average number of CEUs completed by those who were retained were around 81 (the line where the purple meets the green) while the average for those lost was 62.
In the second graph, the number “0” indicates that the member was retained while the number “1” shows that they left the organization. Along the y-axis, we see how many CEUs the member took over the course of the year, with the members broken into 10 equal segments.
On the left, of the members that were retained, 65% of them completed CEUs. On the right, as we get closer to “1” along the x-axis, fewer and fewer members have completed even one.
Clearly, there is a huge correlation between the number of courses that were taken and whether the member left. And although there are other metrics that showed a lack of interest in the association (few logins to website, few comments on blog posts, etc.), this is a clear metric that we suggested the association focus on.
So we know why (or at least part of why) the members are leaving. What do we do now?
This case study is a perfect example of why now is a great time to dive deeper into your member data. Members completing fewers CEUs isn’t just predictive of the coronavirus pandemic. It is likely one of the key predictors for this association, even when things are going well.
But because we know that it is a major predictor now, we can tailor the marketing messages to connect better with members and give us a better chance to retain them.
When we develop an action strategy for an association like this, we help define the people who are making up this potential area for better retention. People who have been unable to complete CEUs during COVID-19 might include:
Of course, you know your membership best, so your expertise is crucial to know which of these messages will land with the most members. But it can help guide you in the right direction as to how to best connect with members who are on the verge of leaving your association.
Yes, things are hard right now. But you can make the best of it by leveraging your data and developing smarter marketing through predictive analytics.
At Tasio, we work with associations every day who are dealing with massive losses due to the coronavirus pandemic and help them create strategies for keeping more members with fewer marketing dollars.
If you want to learn more about our DIY method for creating member loss predictions that really work, download the Tasio Association Retention Playbook: The 5-Step DIY Membership Retention Strategy and Workbook.
[BUTTON: Download the ebook and workbook]
To have one of our Data Specialists connect with you about an in-depth data review, contact us here.