119. SPN
SMS list growth 📲; Build media plans in minutes; and, How to test New Audiences with no history in Google Ads
A very warm welcome to all the new subscribers. I’m thrilled to have you as readers and truly appreciate your feedback and support.
Do you fundraise globally AND deliver a localized donor experience?
How about doing so while not going through the pain of reconfiguring your website?
In my experience it’s only doable with Fundraise Up, the platform built for multi-country fundraising.
Game changer? It is for me.
In this week’s SPN:
Automation: Create fully data-driven media plans and execute them in minutes
EOY focus: Opt-ins and list growth for SMS and email
How to test New Audiences with no history in Google Ads
Jobs that took my fancy this week
Autonomous Platforms
This demo of ViantAI isn’t the most exciting use of AI you’ll see but it feels one of the most profound for the ad industry - an end-to-end media planning and buying tool.
Having watched the demo multiple times this week ViantAI is positioning itself as your digital advertising expert - taking over the difficult, data-intensive work, so advertisers like you and me can focus on strategy direction and creativity.
Right now it does the job of a relatively junior agency exec but you can see how it’ll develop. Then imagine how the platforms will tailor their process to integrate with these tools...
Driving SMS Opt-Ins is Critical
There’s only a few things I care about beyond revenue when it comes to Cyber Week and Giving Tuesday (BFCMGT).
And one of those things is SMS opt-ins from the traffic I drive to an optimized landing page.
That’s because when I think about allocating media dollars I’m thinking about the following:
How can we make SMS opt-ins go up in order to maximize our paid spend beyond just immediate donations?
How can we have the perfect post donation SMS flows which incorporate strong upsells to monthly & cross-sells to product (where applicable) to maximize fundraising revenue, repeat donation rates, and overall retention after the initial donation?
And lastly, how can we get more reviews from happy donors to supercharge our social proof flywheel to improve our brand perception and hopefully make our media dollars more efficient over time?
Although these considerations don’t get reported on as easily as CAC and ROAS numbers do, they’re incredibly important to think through, especially when deploying significant amounts of paid media spend in Q4.
Speaking of which: Any time that you’re spending hundreds of thousands or even millions of dollars per month to drive traffic to your site (which you might be doing for BFCMGT and EOY) you’d better be tracking more metrics than just total fundraising revenue and email opens.
That includes metrics like:
Total site sessions
Average time on site
# of donation add to carts
# of donations started
# of abandoned donations
Bounce rates
# of unique visitors
Percentage of net new sessions
Pages visited per session
Total actions taken (i.e buttons clicked, etc)
CVRs by landing page
But probably the most important outside of revenue is SMS (and email) opt-ins. And now is the time to start asking and gathering them.
The reason I say this is because once you have a Supporter’s SMS number, you have the ability to contact them frequently. And outside of political fundraising season the competition for attention and action is miles less.
Manage the frequency correctly (donor experience) and you can build communication workflows around program/new program updates, emergency asks, unique digital event invites and more, and drive repeat donations over time.
Even if Supporters visit your site and don’t become a donor during your BFCMGT ad blitz, getting their SMS number is crucial for continued marketing and comms.
I don’t think anyone has perfect stats on this, but I still think SMS could drive between 20-30% of all revenue for most Org’s. It’s an enormously under-utilized channel that can contribute tons of money to your success.
And what really matters here is the engagement rate x the size of your respective SMS lists for your Org.
SMS numbers are so foundationally important to your ability to reach donors, engage and convert them, retain, and re-engage them over time that I’d argue they should be the second most important metric you track beyond fundraising revenue and ROAS from paid spend.
Explore: Build Your Tech Stack (link)
Searching for new tools or trying to trim down your tech stack? Play around in this infographic. I add to it most weeks.
Jobs & Opps 🛠️
IRC: Chief Information Officer $239,000 - $280,000
Nuestros Pequeños Hermanos (NPH): Vice President, Donor Engagement
Prostrate Cancer UK: Manager, Sporting Events £40,500 - £44,500
Doctors Without Borders/Médecins Sans Frontières - USA: Senior Digital Marketing Manager, Advertising $101,769 - $152,653
Canadian Red Cross: Senior Manager, Transformational Giving (Major) $73,631 - $107,378
Smithsonian: Assistant Director, Online Engagement and Fundraising $117,962 - $152,354
American Heart Association: National SVP, Technology Solutions
—> Many more jobs can be found on Pledgr.com 🎉
How to Test New Audiences with No History in Google Ads?
Look-a-like audiences (LALs) are the cornerstone of every performance fundraising campaign.
Despite their proven performance in driving revenue, LALs aren’t enough to help Org’s reach new audiences - e.g., they’re useless if you’re expanding to a new geography, or expanding programming to include one more pillar with a new donor profile, or wanting to scale faster by increasing “market share.”
LALs won’t help in these three scenarios. But the process I’ve dropped below should - I use it every time I’m looking at launching new audiences when I’ve no historical data available.
Tightly define the demographics of the audience you’re looking to expand into.
500 donations is a good yard stick to know if the audience is performing better or worse than your core donor base, and whether you can make the economics work between the cost to acquire a donor and their donation value.
With an average website conversion rate of 1% for new donors, you need 50,000 visitors to drive 500 donations. And with a CTR of 1%, you need 5,000,000 impressions to drive that many visitors. The average frequency is around 4 impressions, so you need 1,250,000 individuals. Accounting for up to 2x worse performance than the average – you never know with the new audiences! – the total audience size should be no larger than 2,500,000 people.
In Google Ads - the platform I usually use to test new audiences with the most reach - the 5 main dimensions to define demographics are Age, Gender, Household Income, Parental Status, and Geography. Age is defined in ranges of 10 years, and household income is in increments of 10% relative to the average in the country. With a 2.5 million people target, define the segments as follows:
One gender
One age range
Two ranges of household income
Both options for parental status
As many states as needed to get to the 2.5M, and
The “Unknown” option disabled across all dimensions.
SPN Tip: Geography is tricky –stick to only picking states within one region (Northeast, Southeast, Midwest, Southwest, and West) and either only big cities within those states – or states excluding the big cities. For example, the audience can be 25-34-year-old females in the Southeast region’s 40-60% HHI range, excluding Miami, Atlanta, Charlotte, Nashville, and Jacksonville.
2. Segment based on psychographics.
Once the original 2.5 million target audience is defined, break it into smaller segments of ~150,000 people each. With no data uploads, Google Ads offers three options for segment targeting: detailed demographics, interests (affinity), and current research (in-market and life events).
From Detailed Demographics, separate those with high school degrees from those with bachelor’s and advanced degrees. This will divide the original audience into two. (Other dimensions - number of children, marital status, and homeownership status - are less accurate for targeting and are rarely as impactful).
Skip the Affinity section. Google’s definition of affinity is broad, based on long-term search trends, and isn’t replicable in other platforms.
From the life events section, use the “recently started a new job” dimension, separating the audience further into four segments. This dimension is a good predictor of a recent change in income.
Finally, identify two dimensions loosely related to your programming from the in-market segments. For example, two types of children’s products for an Org supporting children, education products for an Org providing access to education, or health products for an Org funding medical research. This will take the total number of segments to 16, with roughly 150,000 individuals in each segment.
3. Create display campaigns for each of the resulting segments with the same creative assets.
Adding a variety of creative assets to the mix is detrimental to the test. Use one best-performing asset from your current prospecting and retargeting efforts, which creates a total of 32 campaigns:
16 prospecting campaigns, with a frequency of 4 impressions
16 retargeting campaigns separated into the same segments, with a frequency of 2 impressions
4. Launch the campaigns with fixed CPM bidding.
For the initial launch, fixed CPM will provide the quickest insight into segments that will – or will not – work. Over time, an automated setup will be the right way to go, but initially, it would obscure the by-segment insights. Pick the average CPM of your current Prospecting and Retargeting campaigns and use those as the starting point.
5. Aggressively turn off under-performers and scale over-performers.
Every week, for four weeks, turn off two worst-performing prospecting campaigns and two correlating retargeting campaigns. Alongside turning the two campaigns off, increase the budget for the two best-performing campaigns by 25%.
SPN Tip: If a campaign with the increased budget turns up in the worst-performing list at any point, return its budget to the initial figure and delete the next-worst campaign. If it still appears in the list of worst performers a week later, disable it.
Then after the first month, disable half of the original campaign structure, with eight segments remaining. If the cost per acquisition for any of these eight segments is no more than 2x higher than your “core” prospecting campaigns’ acquisition cost (CPA), keep the test running for another month, turning off only one campaign every week instead of two.
At the end of the second month, if campaigns are still performing with a CPA no more than 2x higher than average, include them in your “core” campaign structure, optimizing creative assets and bidding until the new audience hits 500 donations.
Finally, once the campaigns hit 500 donations, the average CPA of the last 50 is a good predictor for the future.
If it’s lower than the average donation value and the RoAS is positive, keep the new audience in the mix.
There you have it. Let me know if you deploy this approach and how you get on. I’m happy to help.
That’s all for today!
If you enjoyed this SPN, please consider sharing with your network. Thank you to those that do.
If a friend sent this to you, get the next edition of SPN by signing up below.
And huge thanks to this Quarter’s sponsor Fundraise Up for creating a new standard for online giving.
Reads From My Week
How Netflix won the streaming wars (FT)
Attention and Memory - learnings from 9000 brand lift studies (Havas)
Here’s what I made of Snap’s new augmented-reality Spectacles (MIT Review)
Programmatic DOOH grows as advertisers move budget from other channels (The Media Leader)
Google’s Ad Business Is Tailor-Made To Be Cut Apart (Bloomberg)
Buried DOJ Evidence Reveals How Google Dealt With The Trade Desk (AdExchanger)
Hiring Trends in Marketing Squeeze Brand Managers (Bain & Co)
Apple AirPods Pro Granted FDA Approval To Serve As Hearing Aids (Techcrunch)
Creating More Value With Your Integrated Sales Tech Stack (BCG)
As political ad spend faucet opens, CTV media stands to be a major winner (Digiday)