Is Your Measurement Missing the Mark?

Most organizations use considerable energy answering one question: Am I spending my resources effectively? Then another question is posed: Is my measurement accurate?

Nonprofit professionals fall into one of two categories:

1. My reporting could be improved, but it is time tested and familiar.
2. I need reporting on every interaction my organization has with a donor.

For the first group, think of a battle-tested metric—cost per donor acquired. Unfortunately, this metric is not as accurate as we lead ourselves to believe. Most consider cost to be dollars spent within a channel (direct mail) over a given timeframe (May acquisition, second quarter, year). What about previous attempts to acquire this donor? Or attempts in other channels?

To the second group, this is very ambitious and most organizations are not at all ready for it. Full attribution is expensive, takes organizational commitment and is hard. Most are not able to take full advantage of attribution even with the right technology, analytics and reporting.

There is hope for both groups. We can attribute more accurately without extreme costs and time.

Assess Where You Are

The starting point is to assess where you are on this scale.
- Use cost-per-dollar raised in each channel based on results within that channel (response, click-through). Over 95 percent of organizations are at this level.
- Use media mix modeling to make budgetary decisions.
- Combine media mix modeling with individual attribution to make real-time decisions.

Understanding today’s approach is an important self-reflection point to ensure you start from the right place.

Take Attribution One Step at a Time, But Don’t Always Move to the Next Step

If using cost per dollar raised by channel, the first step in your attribution journey is test and control. Most organizations can improve spend with a series of well-designed tests. Identify which channels share the most donors (likely email and direct mail) and design test groups of direct mail only, design test groups of email only and coordinate direct mail and email. Run these test panels over three to six months to get the best read. Measure all donations from the individuals in the test and not just the donations received within the tested channels. This test provides a more accurate cost per dollar raised for direct mail and email. From here, add channels to the tests until you cover a large percentage of spend.

Media Mix Attribution

Many sophisticated marketers learn what they need to know through test and control attribution. If you wonder about the sources of unknown donors and spend a considerable amount on less direct channels (TV, print, digital acquisition), then your next step is media mix attribution. Media mix is a powerful analytic tool that determines cost-per-dollar raised by considering marketing spend for all channels and understanding that spends impact on donations. While not individual-level, the analysis can go as deep as your data allows. In some cases, commercial organizations attribute television revenue at the DMA level by day by day-part by channel. Media mix modeling also builds saturation curves that enable scenario planning. You can predict the cost per for the next X dollars spent in a channel and then optimize the spend across channels. Media mix modeling is statistically sophisticated and requires expertise that most organization do not have in house.

Fractional (or Full) Attribution

Only the most robust marketers will ever need to look beyond media mix modeling attribution. Many are naturally curious about the impact of each marketing interaction on each individual, but most would not take full advantage of fractional attribution. The few commercial organizations using this attribution are mostly invested in digital and use machine learning to personalize their marketing, site and mobile engagements based on an individual’s self-driven path.

Still curious? Fractional attribution relies on creating an event stream of every interaction with all known individuals and building models to determine the impact of each interaction. The models identify like individuals who take similar paths. The small differences in those paths provide insight into the impact of each step in the journey on revenue/sales. The most difficult part is collecting the event stream data and identifying the true individuals versus their digital proxies. What’s the good news? The world is moving away from digital proxies (cookies and device IDs) back to a people-based approach. By the time most are ready from full attribution, it will be much easier.

Take Action and Self Assess

The right attribution for your organization opens a new world of more effective spending. The right attribution enables you to raise more money or give money back to the organization to spend on its important mission. Now is the time to assess where you are today, identify the right next step and start spending your marketing dollars more accurately.