What I wish I knew when I was new to machine learning

Including machine learning into your marketing mix is a journey and commitment. As fundraisers, we understand journeys. We build them for our donors all the time. Sometimes we provide a stellar donor touchpoint that surprises and delights our donors along their path with your organization. Sometimes me miss the mark and don’t quite connect the way we intended. But we learn and grow with every constituent touchpoint. As long as you’re engaging with them, your donors—more often than not—will continue to follow along with your non-profit’s journey.


Here’s a few helpful tips that I wish I had known before I began my journey with Fundmetric. May they help you along your way to data enlightenment.


1. Pace yourself for the long haul.

With machine learning, you might hit a home run right away with your very first effort. The next few efforts might not be so impactful. What you’ll find is that over several campaigns the machine gets smarter and smarter. Before you know it, you’re on a steady growth path to meaningful engagements with your constituents that translates into donors that stick with you longer, give more and more often. Plan on testing several campaigns and track your engagement metrics over the entire testing period. You’ll see a steady increase in engagement over time, which likely means you’re constituents are finding value in your communications.


2. Identify what success looks like at the start. 

As fundraisers, we naturally have a bit of a competitive streak within us and we want instant and unquestionable success with every campaign. Strengthening the machine learning model will take time and you might not hit it out of the park with your first few efforts. Fundmetric was very upfront about that when I signed on and this is common for new subscribers. 


Understanding this learning curve and identifying what success will look like along your journey before you begin can help manage expectations. You’re in it for the long haul, not just a quick and possibly non-replicable win. Instead of focusing on the immediate, set milestones that make sense for your fundraising program and assess your KPIs periodically. Remember, revenue is only one indicator of success. Focusing on LTV and sharing regular and meaningful engagements with your donors will naturally lead to increased revenue.  


3. Don’t go at it alone. 

If you have multiple teammates or colleagues from other departments that could benefit from Fundmetric, then get everyone to assist in the implementation process right from the beginning. You can still (and should) assign one or two people to lead the project, but getting every potential user’s input as you’re setting up shop will help to build a better foundation for the machine learning process. This will culminate in successes that come faster for everyone. 


If you’re at a larger organization like mine, you likely suffer from what I call the ‘silo effect’— where one team doesn’t always know what the other is doing. Working together to onboard Fundmetric will help facilitate communications between all teams and colleagues involved. You will all learn the ins and outs of Fundmetric together, building synergy, and best of all, you’ll test Fundmetric with multiple levels of your organization simultaneously. This will exponentially feed the machine learning pipeline and maximize your ROI.


4. The predictions you get with Fundmetric are more advanced than competitors. 

Not all artificial intelligence platforms are created equal. Fundmetric was the natural choice for my fundraising program because it was the only system that could take into account behavioural information from our fundraising channels (direct mail, digital, events, peer-to-peer) and our e-commerce channels to build a wholistic prediction model that we could use across all channels. 


This was extremely important for my needs. Competitors of Fundmetric are good at helping your team manage their donor prospects by recommending who your donor managers should reach out to next. But that’s where it ends with them. And if your data doesn’t fit into their one-size-fits-all algorithm, then it’s all for naught. Or if you would like to add prediction-based insight to other initiatives, you’re out of luck.


Unfortunately for some non-profits, they have spent the time and money to implement a new platform , only to outgrow it within months. With Fundmetric, you can set up prediction cards for any business need you may have. You’ll go above and beyond what other platforms offer almost immediately.


5. Fundmetric could become your only email deployment system. 

That wasn’t something we had planned on happening when we embarked on our data enlightenment journey. We just assumed that we would run Fundmetric alongside our existing email deployment system and use the machine learning to inform our marketing strategy.


Turns out that Fundmetric was able to do things we couldn’t using our old email platform. Fundmetric’s donor portal featured a smarter design and looked better in a mobile-responsive setting than our previous donor portal. The portal is bilingual and allows for a multitude of giving options (one-time, monthly, annual, etc.) that could turned on or off with ease. 


Another added bonus—being able to use donor-specific suggested gift amounts on donor portals based on machine-informed predictions. This was a functionality that our existing email deployment system claimed it could do, but we always experienced issues and generated errors. With Fundmetric, there will be no more asking the same amount from each recipient of your digital campaigns. Instead, Fundmetric recommends asks that are appropriate based on donor behaviour, giving history, social media indicators and so much more. 


Lastly, the email builder was so simple to use that even I could create emails. Yes, that’s saying something! I’m not the most technologically savvy knife in the drawer. But building and designing emails was literally as simple as 1, 2, 3. 


With all these benefits, you could easily replace your existing email deployment system with Fundmetric, saving the costs of multiple systems while getting the value add of a superior machine learning platform. 


6. Assign meaningful data tags to your data before it goes to Fundmetric.

Although naming your campaigns by their traditional names is great… Year-End Renewal, House campaign, Prospect 1 and Annual Appeal mean nothing to few more than the folks that track and report on their revenue. Taking some time to assign meaningful tags and values to your campaigns will help build an a predictive model that does more than just tell you which donors need your attention. Consider ways you can provide the machine with data that means more in the long run. For instance, something as simple as changing the naming convention of your campaigns can help you get the most out of Fundmetric faster.  


Renaming campaigns based on the year and month they deploy, the channel by which they deployed, the type of ask (renewal, second gift, monthly conversion, etc.), the specific program(s) the funds will support and the audience it went to can help accelerate the machine learning curve. So perhaps your year-end renewal could now be: 2020 December renewal mailing to mid-level donors restricted to XYZ program. Or your digital lead-generation campaign could be: 2020 October monthly donor acquisition digital campaign to non-donors area of greatest need


By doing this, you’re now identifying segmentation factors that Fundmetric can use to build enhanced prediction models. Over time, Fundmetric will be able to tell which donors are likely to give at what time of year via their preferred channel and to what programs they’re likely to give. Yup! Mind blown!


7. The level of customer service at Fundmetric is unparalleled. 

How many times have we been promised great customer service after you sign on with a vendor but after you’re on board, you become nothing but a number? That’s the Fundmetric difference. They are a true partner with you and your cause. Your successes are theirs. Your losses are just as heartfelt. So as true partners, Fundmetric steps up to provide not only a dedicated Account Manager, but access to a team of data analysts, a data scientist and consultants, like me, to answer your questions, look at the intricacies of your organization’s specific needs to strategize  and grow together. 


I hope you find these tips helpful. I think most of all, just take the time to enjoy the journey with Fundmetric. We’re here with you every step of the way. 

Bio: Kristopher has over 18 years of marketing experience in both Canada and the USA and 8 years experience in fundraising for Canadian charities. With an emphasis on multi-channel direct marketing, Kristopher has managed over $7 million dollars in annual donations integrating direct mail, digital including predictive modelling, face-to-face and telemarketing strategies to drive growth and lifelong donor journeys. 

“The concept of digital fundraising today must include predictive modelling/machine learning. Including machine learning in the mix ensures that you’re driving down your cost of funds raised while ensuring that no donor feels overlooked because you’re providing meaningful, personalized stewardship touch points at the right time in their donor journey.”
-Kristopher Gallub, Fundmetric Fundraising Liaison