You can find the article here.

Today’s experiment

WildCare Bay Area is a wildlife animal hospital in Marin.  WildCare’s staff wrote an email about baby pigeons, a new batch of which was in the hospital.   The subject line of the email was A/B tested.  The actual email was a series of questions and a photo of the birds in our hospital.  To get the answer, you had to click through to the website.

The contenders

Original subject line (Group A): “What ARE those creatures?” VS  Variant subject line (Group B): “Guess what kind of animal we are!”

The results

Thanks to Usereffect.com for their A/B calculator

The Safdar Group is a boutique digital agency serving causes and non-profits. We have a deep commitment to analytics, both in helping organizations roll out analytics, but also in providing analytically-driven traditional digital agency services to our clients. We use analytics in how we do everything, from write headlines and subject lines, to choosing layout and colors.  The A/B test diary contains results of A/B tests from actual client work, with permission.   

To learn more about the services of The Safdar Group, contact us at +1.415.683.7526.

There are a number of interesting analysis techniques that you should run over a long-term sample of your email marketing work that will tell you valuables things about their effectiveness.   Two of the ones I’ve performed for clients are “email shelf life” and “subject element correlation” analysis.  Today I’ll talk about “shelf life”.

Email shelf life analysis

Do you know how long your email is really productive?   How you time your emails and how long you can expect them to perform is an important question.  Can you really go two weeks between emails and have them still generate clicks?   One of the types of analyses I perform for my nonprofit clients is to look at the activity generated by their email and on their website over time, and see exactly how long people still interact with it.

I typically measure activity, in particular clicks, in hours from delivery completion.   Here’s an example of a graph of clicks on an email over time from such a study:

This graph shows that the most activity came in the seven hours after it was delivered.  At that point night fell and activity picked up again the next day, and so on with diminishing returns.   By my eyeball, this email was still performing 100 hours out.   However you need better statistics than eyeballing a graph, so instead I created a metric called “90% performance of clicks“.   In other words, if you measure an email far past it’s useful life, exactly how many hours does it take to earn 90% of it’s lifetime of clicks?   For this email it was 3.9 days, or 95 hours.   Others I studied for this same client generated 90% of their performance in roughly 5 days.

This leads to two important questions of frequency and timing for anyone trying to maximize their email results. Notably:

1. Does my audience click on my emails and respond on the weekend?  If they don’t and my emails have a 90% performance rate of four days, then I pretty much need to send all my emails no later than Tuesday morning of any given week.
2. Am I waiting too long between emails?  If the 90% performance rate of your emails is 5 days, but you only send an email once every two weeks, why would your audience give you any money, or do anything for you during the off week?

At all times, you should be either conducting experiments on your email campaign, or studying the results.  If you’re ever not doing either, you’re wasting a perfectly good opportunity to get a higher yield (dollars, volunteers, advocacy actions) from your list.

And once about every 6-8 months, you should be doing a long-term study of your email program for questions like these.

In my ongoing quest to fully install Google Analytics on Convio’s platform, I have sadly found another bug, this one having to do with cross-domain Google Analytics connections.

One of the trickiest things to do in analytics deployment is to correctly track a visitor across a number of domains where you have pages, like www.savethewhales.com, store.savethewhales.com and secure3.convio.net.

Many convio clients actually have a setup very much like this.  They have their main site either on the Convio platform or on their own platform, but their default donation pages are on Convio’s SSL-secured hostnames at secure2.convio.net.

If you know anything about cookies, which is how Google Analytics tracks visitors, you know this won’t work.  The cookies which uniquely identify you every time you return to www.savethewhales.com aren’t shared with secure3.convio.net.   So you look like 2 different people.  What’s worse is that the data about how you arrived at the site, which is crucial to understanding where your conversions are coming from, will be lost.

For example, a standard report you want to run on everyone that donates is “how did they get to my site, and how many times did they come to my site before they gave?”  All this data is lost if you don’t carefully pass the cookies from www.savethewhales.org to secure2.convio.net.

How Google Analytics passes your identity from domain to domain

GA has a solution for this, they put a copy of your cookies into the URL and at the other end, they set them.  So, if when surfing www.savethewhales.org, GA says you’re visitor #66666, then when you click the link to Donate that would normally take you to

https://secure2.convio.net/donationform.html

instead GA appends your id to it, like this:

https://secure2.convio.net/donationform.html?analytics_id=66666

So what’s the problem?   Convio’s Donation application (and possibly others) appear to reload themselves and silently delete the Google Analytics cookie in the process of setting a URL parameter idb.   So instead of landing on

https://secure2.convio.net/donationform.html?analytics_id=66666&idb=67654

you land on

https://secure2.convio.net/donationform.html?idb=67654

Here’s an actual example of a URL progression from one of my clients:

Original page:

http://www.wildcarebayarea.org/site/PageServer

User clicks on a donation form with the following URL

https://secure2.convio.net/wc/site/Donation2?df_id=1360&DONATION_LEVEL_ID_SELECTED=1&1360.donation=root

Google Analytics appends the cookies and turns the URL into this:

https://secure2.convio.net/wc/site/Donation2df_id=1360&DONATION_LEVEL_ID_SELECTED=1&1360.donation=root&__utma=1.419819392.1345758089.1345758089.1345758089.1&__utmb=1.1.10.1345758089&__utmc=1&__utmx=-&__utmz=1.1345758089.1.1.utmcsr=(direct)|utmccn=(direct)|utmcmd=(none)&__utmv=-&__utmk=77686036

All those parameters that looks like __utm are Google Analytics cookies.  Then the bad thing happened.  Upon arrival at secure2.convio.net, I used a sniffer to confirm that upon arrival, the browser was redirected to:

https://secure2.convio.net/wc/site/Donation2?idb=404683257&DONATION_LEVEL_ID_SELECTED=1&df_id=1360&1360.donation=form1&JServSessionIdr004=t22c5n4tw4.app207b

Whoa, where’s my Google Analytics cookies?  Where’s my campaign conversion data?  Where’s all my valuable marketing info to tell me what marketing techniques drove my people to give?  GONE.

There are a couple of possible fixes to this:

• Assuming all I need here is the idb number of the user, it would be nice to know exactly how to get that and insert it into the URL so a redirect wasn’t spawned.  I tried using idb=[[S76:idb]], but that didn’t work.  It just generated idb=0, and I still got a reload.
• The better fix is that when applications like Donation2 perform a reload/redirect, they should preserve all the URL parameters, including the ones they don’t know about.

Fix:

Noah Cooper, a developer at Convio, has confirmed that this is a bug (#63189)

As a workaround, he explains that you can link directly to the first page of a form and avoid the redirect.  Here’s his explanation of why you should not include “1360.donation=root” in your URL when linking to a donation page:

Donation forms can either have a splash page, or not (most do not).The query string 1360.donation=root tells our product "I'm not sure whether this donation form has a splash page or not, I need you to figure that out and take the user to the right URL".

Since this form does not have a splash page, requests to https://secure2.convio.net/wc/site/Donation2?df_id=1360&DONATION_LEVEL_ID_SELECTED=1&1360.donation=root result in a server-side redirect to https://secure2.convio.net/wc/site/Donation2?df_id=1360&DONATION_LEVEL_ID_SELECTED=1&1360.donation=form1, the first step of the donation form. That server-side redirect does in fact drop any query strings that were appended to the original URL.

If, however, you change the URL so that it includes 1360.donation=form1, the server-side redirect isn't needed, thus the query strings are retained.

This will work for probably everyone for the moment as a workaround.  However the destruction of the GA analytics URL query parameters is still being treated as a bug and (presumably) slated to be fixed.

[Updated 8/19/2012: I've found yet two more bugs that have me pulling my hair out.  They are documented at the bottom.  -S]

Convio’s CMS and transaction processor offers integration with Google Analytics.  This is great, as there are distinct elements of the donation or e-retail experience that are difficult to collect data on without their assistance.  However it seems like the Google Analytics integration was implemented by someone who followed the directions but didn’t necessarily pull any data to analyze.  There are a number of fatal errors, several of them documented.

I’ve worked on a number of nonprofit websites based on Convio, and I’m always being hired to fix what is broken out of the box.  In the interest of actually making the situation better, I offer three five six issues and suggested fixes that would allow people to actually use the data in a meaningful way.

Correctly implement redirects on secure*.convio.net. (Convio bug #63189)

This one is so complicated it requires its own blog post, but here’s a short summary.   When you have your main website somewhere like www.savethewhales.com and your donation pages on Convio at secure3.convio.net, Google Analytics will not work right out of the box.  You need to trick out the links to secure3.convio.net with special code to synchronize your web visitor’s id across from your domain to the Convio domain.

I’ve discovered that Convio’s web app on secure*.convio.net is deleting the Google Analytics cookies from the URL and reloading the page in order to add paramters like “idb” to the URL.  Unless there’s a way around this, pretty much every Google Analytics implementation that uses secure*.convio.net is broken.

Fix: Tweak applications like Donation2 and Advocacy to not delete the Google Analytics cookies when they do a redirect to themselves (Convio bug #63189)

Properly implement the Google Analytics asynchronous tagging (Convio bug #62673)

Historically, the donation, advocacy, e-retail and other transactional parts of convio’s system are built using the legacy Google Analytics tagging (the ”synchronous” code, as opposed to the more recent and improved ”asynchronous” code).   The old, synchronous code was placed at the bottom of the web page, and in some cases it didn’t run because the page had to fully load.   This resulted in a loss of data.

Last winter, Convio implemented the new, asynchronous code, but left it at the bottom of the page, instead of following Google’s directions and placing it in the <head> section of the document.

Fix: Move the standard appearance of the GA code to the end of the <head> section of pages and closely follow the GA recommendations.

Add cross-domain cookie passing code to e-retail, donation, and advocacy Google Analytics tags

Many Convio cms-based sites are built like this:

1. Main website appears on branded domain name (www.stopshootingsealions.org), which link to…
2. Donation and advocacy pages appear on https://secure3.convio.net/

If you know a little bit about Google Analytics, you know that this setup doesn’t work without changing the links on stopshootingsealions.org and configuring the GA code on Convio’s secure pages to appropriately link the web visitor across domains.   Why is this important?

Imagine if you google “are people shooting sea lions?” and end up finding stopshootingsealions.org and giving them a donation.  By default, the fact that you’re a Google-search driven web visitor will not be recorded as the source of that donation. In fact if you run a “source of conversions” report in Google Analytics, it will look like all your donors came from one website, “stopshootingsealions.org”.   Unhelpful.

The default, out of the box Convio template for Google Analytics code doesn’t include the linker code on the Donation, Advocacy, or e-retail pages, and there’s no way for a Convio admin to alter the GA template to have it synchronize the cookies.

How do I know this?  Because I have a tidy little consulting business implementing cross-domain cookie synchronization for Convio clients.  Though it’s probably not in my personal best interest, it’s in the best interest of nonprofits everywhere if Convio fixes this.

Fix: Add GA Linker code to Convio GA template on all secured Convio-hosted pages, or allow Convio admins to change their GA tags template for these pages.

Openly document the GA tags template

Despite everything above, I really like the Convio GA implementation.  The Convio implementation of the “form hierarchy” pseudo URLs that makes reporting so much easier.

Examples of this include: “/funnel/Donation2/<name of form>/step1″ or ”/funnel/Donation2/<name of form>/complete”. This sort of naming convention that Convio uses makes it easy for web analysts to setup
goals, funnels, custom reports, etc.

Since it’s non-trivial to convert this to use the asynchronous code, most convio clients with real analytics needs I know are still using legacy items on their transactional pages. You can get away with this for a long time until
you start using a product like Optimizely that really, really, wants to report data back into GA using only the new asynchronous code.

Somewhere, accessible by the Java code that runs the Convio system, is a Javascript template that contains the Google Analytics code for e-retail, Advocacy, and Donation pages.  It probably gets (or could get) all the variable data like “transaction amount” from Convio’s S-Tags.

Like many other parts of the Convio system, it needs to be opened.  Others who have specific needs for their GA code that Convio hasn’t implemented (and that will be anyone who’s seriously using GA for custom segmentation) need to be able to customize it at their own risk.  Ideally it would be a template editable by the admin.  Less ideally, it would be published as a reference implementation so that others can implement their own forked versions.

Fix: Allow admins to edit the GA Javascript template, or publish a reference implementation for others to fork and customize.

Fix the bug that causes Advocacy funnel reporting to fail (Convio bug #63113)

Convio’s GA integration touts the ability to report all the stages of advocacy participation in a funnel, from the Splash page through completion.  That’s great, except that it doesn’t record the first page of the process (the splash page) due to a known bug recorded by Convio.

Fix: Correct defect outlined in existing bug report.

Disallow “second pageviews” in Convio’s Google Analytics tracking (Convio bug #59403)

If you are using funnel tracking in Convio, you may have noticed something funny about the code it inserts.   It looks like this:

Instantiating a tracker and then making another analytics call is how Convio is generating a virtual pageview for funnel tracking.  Except there’s a problem.  Your analytics will now tell you that twice as many people went through your advocacy or donation process as actually did.   And your pageviews will be wrong.   And worse yet, both of these pageviews have separate Google Analytics visitor ids. I’ll bet it will also interfere with conversion source tracking.

The fix to this is really easy. All you have to do is call the original trackPageView call with the url you want to record in your analytics, like this:

_gaq.push(['_trackPageview','/funnel/Advocacy/name of advocacy action/UserAction']);

The resulting code would look like this:

Fix: Correct defect as documented in Convio’s bug database.

Noah Cooper at Convio has heard my points, and been very amenable, but he’s only one man.  Give him a little help by voting these up on the Convio Community in this post.

This is what 'Ricky' at Optimizely support looks like. At least to me.

When Google announced they were killing off their classic Google Website Optimizer product this year and implementing Google Content Experiments right inside Google Analytics, I was super excited.  Tighter integration is great!  But there’s been a lot of hand-wringing since then.  I actually tried creating an experiment with it and had some difficulty with getting it setup right.  There are help pages on Google, but very little in the way of other community support.   I only found one post from someone else having trouble with GCE, which did not bode well.

So instead I decided to try one of the competitors, optimizely.  There’s a lot of additional features to optimizely that are really attractive, though I will always be a Google Analytics devotee, I decided I could stray a bit.

Lo and behold, I had trouble with that product too, and unlike GCE, it’s not free!   Interesting though, they were all over me like hipsters on a food truck.  As soon as I registered I got a call from my account person.  When I explained I had trouble, the account guy (not a support guy) asked me to send an email and he’d make sure a case got opened for it.   I ended up posting a note one day right before 5pm and got a response about 20 minutes later.

From there on I had an ongoing email debugging session with a support gentleman named “Ricky” at optimizely who would have responses to my emails in under 4 hours every time, and most of the time in an hour or less.   I even took a day off when I got busy with another project, and didn’t do the debugging task I had to do to move the project forward.   A day later when I did?  I had an email 90 minutes later.

If you’ve never done support you probably don’t understand how hard it is to maintain a full caseload across multiple days and keep that kind of response time.   I’m tres impressed.

I may never completely give up Google Content Experiments, but optimizely is getting my monthly subscription fee, for sure.

Photo credit: Gears of War gunsaw

Self-denial is an awesomely powerful management tool.

Your goals and KPI’s are business focused

Metrics are bits of data, Key Performance Indicators indicate the success of your business.  Here’s how to tell if you have a good KPI: if your KPI directly impacts your ability to continue to get a paycheck, it’s a good KPI.   Good KPI’s are things like money, volunteers, visitors to your facility.  If your digital marketing makes these things go up, you will probably continue to get a paycheck.

Less desirable KPIs are things like email addresses, likes and follows on social media.   If these things go up, you may or may not be able to make money from them, and you may not be able to continue receiving a paycheck.

One staffer pulls your metrics every week on the same day and talks through them with your colleagues

The kiss of death for any non-profit analytics program is to have to share a spreadsheet between two different people who pull numbers from different sources.  Inevitably it takes most of a week to pull the numbers, and then when you get around to studying them, the data you’re looking at is two weeks old and your ability to react is slower.

I recommend for my clients that they end their weeks on a Sunday, pull their metrics on a Monday, and then meet to discuss them on a Tuesday, no earlier than 10:30am.   This way, if you’re out of the office on a Monday, there’s still time to pull them on Tuesday.

Your KPIs are all actionable

There’s a variation on an old saying that goes, “You can’t manage what you can’t measure”.   This implies as well, “You shouldn’t measure what you can’t manage.”   Nothing makes a non-profit analytics program irrelevant more quickly than looking at a number that varies week after week that you can’t do anything about.  After a while, smart people usually say, “Why is it important for us to study this number?”  And then they stop.

If you don’t have a very clear plan for what you will do to improve a degrading KPI, then perhaps it’s something that shouldn’t be a weekly KPI.  Perhaps you should pull it monthly (or even less often), because influencing it is going to require large, strategic changes, instead of the small tactical ones that you can do from week to week.

You have very few KPIs, and therefore don’t need a complex, full screen dashboard

Having too much data to pull makes the process take a long time, and can impact your ability to pull them on a regular basis.   If you feel that you need a large, full screen dashboard to present your KPIs, you’re probably doing it wrong by looking at too many numbers.  Having four or five for your executive team is a lot.

You occasionally dive down into channels and study their metrics, but you don’t confuse them with KPIs

Your open rate, your click thru rate, and your unsubscribe rate are all indicative of the health of your email marketing channel efforts, but they are not indicative of your business success.  The amount of money an average email raises per delivered piece, however, is most certainly a KPI of your business.

This doesn’t mean you don’t study your email marketing program with analytical rigor, it just means you don’t confuse email-channel-metrics with business performance goals.

You keep your data in a very accessible format, such as Excel

Armed with nothing more than an Excel spreadsheet, you can conquer mountains.  ”But why should I pull that data out of my analytics program and put it into Excel?”  Because:

• Analytics tools don’t keep data forever (Google Analytics drops your data after 24 months).
• Graphing things is really easy when your data is already in Excel.
• Excel is the lowest common denominator of office tools so everyone can view your data.
• Excel also has some pretty bad ass statistical analysis tools, like correlation and standard deviation.
• You won’t have to worry about your data changing out of from under you if something in the cloud changes.

Here’s to hoping you can find at least one of these flaws in your nonprofit analytics program, and change it for the better!

Earlier this year I signed up to be a donor to the Bone Marrow Registry at BeTheMatch.org.  I’ve been thinking about this for a while, but a series of interactions pushed me over the edge. They were:

•    A public appeal for people to sign up to find a match for a child I didn’t know (online/social channel);
•    A personal friend who needed a match to cure his long-standing cancer (personal connection);
•    A segment on a morning talk show about how easy it is (television!).

(Multi-channel marketing nerds will note that it’s nearly impossible to capture this stream, and yet re-affirms everything we know about multi-channel marketing warming up a prospect and converting them.)

The two step technique was simple to execute, here’s how BeTheMatch did it.

1. They emailed me, pointing out that something small that I did was incredibly important to them and asked me to put it on my Facebook Timeline as a “life event”

Honestly I don’t think about my marrow donation much day to day, just when it comes up in conversation or when I see an ad of some sort.  I felt slightly emotional when they emailed me suggesting it was a life event. For someone else, if we match, it surely will be.  Once that realization clicked, I was hooked.

2. Provided me a catchy graphic to use in my timeline

BeTheMatch sent this one out along with these instructions.

If I was designing it I would have made the URL a little bigger, but hey, nobody’s perfect.  Also they encouraged me to write my own message.  I did, but my friend actually put the URL in there, something they forgot to tell me to do.  Look at these two examples and wonder which one is more likely to drive a click to your website.

In this one I wrote a passionate message off-the-cuff, but forgot to include a URL back to the registry where you can sign up to be a donor.

My friend saw my note but had the presence of mind to add a URL to the registry.  Anyone that sees this will have a quick click over to a conversion.

3. The Impact

The impact of this was huge, though.  When I added it to my timeline, it hit my news stream and ALL my friends saw it.  It was a Facebook Life Event, so you can be guaranteed more of them saw it than my recent post about how I shot a really good score at the archery range or about my great camping trip.

A friend who also signed up for the registry saw what I had done and presumably got the same email and did it also.

Secondly, it went into my Timeline of Life Events, alongside the birth of my two children, my marriage, and a half a dozen other related items.  That’s some pretty premier neighborhood real estate.  If you’re looking at my Life Event Timeline, that’s a lot of great coverage.

How Would You Measure This?

The challenge of this strategy is that it’s difficult to measure, but there are a few techniques that can work well for your organization.

1. Watch for fan acquisition spikes or Facebook interactivity

You should have a daily recording of fan acquisition as well as Facebook activity and have a pretty good idea of what constitutes a data spike.  Monitor it while the campaign is running and look for increased mentions.

2. Monitor Facebook generated traffic

Watch your baseline of traffic coming in from Facebook.  You’ll need a long baseline to rule out seasonal variations.  This would only record people who happened to follow a link from Facebook though.  If someone saw it and thought, “Hey, that looks cool!” and fired up another browser window and typed in your URL, you wouldn’t see a referral from Facebook, you’d see “direct” traffic.

3. Watch for a spike in email signups and donation completion rate

You should already have a pretty good sense of what percentage of your traffic signs up for your email list, in the form of “For every 1,000 visitors we get, 25 of them sign up for email”.   If this subscription rate spikes, or the completion rate on your forms go up, that’s a sign you’ve got motivated prospects coming in.  If you’re running no other promotion, then this probably due to this sort of campaign.

Of course, you can also ask your donors what marketing touches affected their decision.

Best Practices

There’s a couple of best practices you can take away from this example:

1. Provide your users an eye-catching graphic, or better yet, a few to choose from.  The 300×240 size is pretty good.  Include a sizable URL in the graphic.  If you are in Animal Welfare and pets, consider letting your users add their own photo to a template you provide to personalize it.
2. Encourage your users to write a personal note for the timeline and don’t let them forget to include a URL to your page or to tag you in the posting.
3. Identify your baseline Facebook activity and website clicks from Facebook.  Look for a lift due to your Timeline campaign.

*Shabbir Imber Safdar is the Chief Data Nerd at The Safdar Analytics Group, a San Francisco-based boutique agency that specializes in digital measurement and innovative online cause campaigns.  He’s a certified digital analyst by the Digital Analytics Association.  Occasionally he gets paid to make earthquake videos and destroy fishtanks and wineracks

One of the questions that I struggle with when thinking about the measurement strategy is always choosing exactly what to measure and how to suggest they handle those measurements.   As Avinash Kaushik is fond of saying, “You are what you measure“.   I often  worry that my clients are measuring things that don’t matter to their bottom line, or are entirely out of their control.   This is what is referred to as “web reporting”, “data reporting”, or even more derisively as “data puke”.

You often end up with bad data because a HiPPO asks for a metric they’ve heard someone else talk about, and then it becomes a Key Performance Indicator (KPI) because they are a HiPPO.

A metric is a piece of very generic data: how many visits came to the website, how many page impressions you had, how many people are on your email list.   However a Key Performance Indicator is a metric with two important features:

1. it has a direct, bottom line impact on the success of the business; and
2. it is analyzable and actionable.

This last statement is crucial.  It is possible to have a metric that has a bottom line impact on the success of your business that isn’t actionable.  That’s a really important metric, but it’s not a KPI.

For example if you’re a farmer and your fields need one inch of rain a week, the amount of rain you get is a really important metric, but it’s not a KPI.  When there’s no rain and you need it, you can’t “make rain”. You can water your fields, but once you do that, you realize that “rainfall” isn’t the right KPI.   “Cost of water used” this week is probably the right KPI.

We have to choose our KPIs carefully, because as each week goes by, we will spend a lot of time studying their variance, and presumably changing our tactics to account for their rise and fall.

Your weekly analysis session with your team should look like this:

1. Review of the data from the previous week, starting with KPIs.
2. Discussion of variance of KPIs.  Did our KPIs rise or fall?  Was this in response to something we did?   Is something we’re doing not working anymore?   Is there a seasonal impact that we can confirm from long-term baseline data?  Did something we do last week fail to move a KPI that should have changed?
3. Resolution of changes in tactics for coming week.
4. Discussion of longer term strategy impacts of results.

The best social media explanation I've seen. Credit

to figure out if all the time spent on Facebook and twitter is actually contributing to the fundraising results of the organization.

Social media is clearly a part of crafting the public image of the organization, and as such, it is important.  But lots of things are important.  If you can optimize it to raise money, then you can make that part of your job, and your own job, really important to the organization.

There are three obvious ways you can measure the contribution of your social media activity to your nonprofit’s bottom line, and the tools now exist for every nonprofit to do them for free.

1. ON-PLATFORM GIFTS: You can measure donations given on a social media platform, such as with Causes or some other social-media-native donation processing application.
2. LAST INTERACTION CONVERSION: You can measure when someone in social media clicks over to your website and makes a donation.
3. ASSISTED CONVERSION: You can measure how often social media drives someone to visit your website in the thirty days before they make a donation.

I’m going to show you how to quickly generate a report that lets you report on #2 and #3 today.  If you’ve already got conversions configured for gifts in Google Analytics, odds are good this data is in GA right now, just waiting for you to pull it out.

How to measure social media-driven conversions

Note that you have to have donation thank you pages marked as conversion events in GA.  If you don’t, then stop and go do that.  Wait a week for data to start flowing then return to these instructions.

Step 1: Go to the Assisted Conversions report in GA, it’s under Conversions->Multi-Channel funnels->Assisted Conversions:

Step 2: Look at your last interaction conversions and assisted conversions:

This report shows two really important numbers:

LAST INTERACTION CONVERSIONS: This is what you get when someone follows from a piece of marketing, comes to your website, and makes a gift.  In this example 25 gifts were made by people who clicked in from organic search results, came to the website, and gave.    In this same example, social media drove 2 people to come to the website and give in the same visit session.

ASSISTED CONVERSIONS: This is the number of conversions that received an “assist” from that channel.  In the example listed here, 11 conversions were preceded in the last 30 days by visits from Organic search results.   In this same example, not a single one of the conversions/gifts during this time period was preceded by a visit from social media.

It’s a sign of a good analyst that you learn something from positive and negative results.    So what should you learn from this big ZERO?

The social media audience that belongs to this nonprofit is coming to the website and giving only in small amounts.  More than that, they are not visiting the website in the thirty days before they give a gift through social media-driven links.  This social media community probably is not giving through the website.

This may be on purpose, but odds are you probably want your social media community to visit your website on occasion and give every once in a while.   In the wake of this, I would start examining exactly what your social media strategy is, how often and how effectively you are pushing people to come to your website for great content, and when you put out an ask to that community.