Welcome to Part 4 of the Ladder 2017 Marketing Plan: an audit of our live email marketing campaigns! Below we’ll go through our current campaigns, looking at performance across MailChimp and Intercom and auditing our newsletter, sales prospecting drip, and CRM messaging.
Email marketing is still an incredible tool for driving performance. It’s the number one lever when it comes to increasing retention, and is particularly useful for getting new users to that ‘aha’ moment as soon as possible.
According to Campaign Monitor email has the highest return of any marketing channel; a whopping $38 in return for every $1 spent.
We don’t necessarily need access to a client’s email platforms to do some basic auditing of their setup. We can simply sign up and see what emails we get!
It’s important you use a unique email address each time you test, because some mail system will de-duplicate emails already signed up. However, we want to actually receive the emails in our inbox… so how do we test multiple times without having to register 50 gmail accounts?
Good news – Gmail has a feature which lets you create a unique address, but still get the email in your inbox. You simple add a ‘+’ before the ‘@’ sign, and type whatever you want in between the ‘+’ and ‘@’ to make it a new email. It’ll still deliver to the same inbox, but the form will think it’s a unique email address.
And we get our email within a few minutes.
Over the next few days, weeks, and months, we’ll get other emails that we can critique. I highly recommend you sign up for your own / your client’s emails because it does give you good ideas of what to fix or test next.
However, as you’ve probably already guessed, this isn’t an ideal approach. Who can afford to wait weeks or months to audit an email sequence?
Some emails are only triggered based on specific actions or only are sent to certain customer segments; if you never take that action or fit into that segment, you’ll never see that email. Even for the emails you do get, you won’t have any data on how well those emails perform.
Email software is probably the busiest space in marketing tech, with multiple billion-dollar platforms all competing against each other. At Ladder, we’ve used MailChimp and Intercom the most, so that’s what we’ll walk through. Don’t worry if you don’t use them — the learnings we list out should apply to any good platform.
To start, we’ll cover the three major types of email campaign: broadcast, drip sequence, and behavioral.
The first type of email a small business typically sets up, is a newsletter. This type of email isn’t personalized; it goes out to everyone, usually once a week. Digging into MailChimp Campaigns, I can find Ladder’s latest broadcast email:
This template is actually great for a number of reasons. It uses large enough font, it’s a very clean design, that also works on mobile. It also has a tailored call to action for the main feature, but gives the user additional links to click if that isn’t interesting. We have the Ladder logo at the top, and some title text that speaks to the main value proposition of Ladder — ‘grow your business without the guesswork’.
The main thing I like about this email, however, is that it doesn’t make the mistake I see most marketers making — it doesn’t over-rely on images. Most marketers come from a branding background, so they naturally want the email to look as pretty as possible. The result is a really nice-looking email, heavily reliant on imagery.
They forget that the nice imagery they chose while browsing on their high-end laptop on super-fast broadband in the office actually loads incredibly slowly for the tired, busy consumer browsing on a small phone screen on 4G internet on the subway. This breaks the whole design and just pisses the user off, hurting conversion rate and increasing unsubscribe rate.
I’ve seen this time and time again, and whenever I’ve tested it, plain-text emails have won out over nicely designed image emails every time. So I’m glad we’re not making that mistake.
Let’s scroll down and take a further look:
One thing to notice, is that we have an address at the bottom (a legal requirement) and we make it easy for users to unsubscribe. Both of these elements are very important in improving trust. What we don’t want is for users to get frustrated and click ‘spam’; if too many people do this it can hurt our deliverability.
It’s also great that we have a call to action at the bottom. This way, we don’t have to be overly sales-y in the email itself, and can naturally start to pick up new clients as they browse and like what they see. The formatting looks a little bit off; we should try and test something new here.
One more thing that we shouldn’t forget; the subject line.
“Marketing Plan and Chill? ?”
This is obviously niche to our audience but building an effective marketing plan is on the top of every marketer and business owner’s mind. The combination of this and a pop culture reference (Netflix and Chill) is probably what go so many people interested in opening this email (40%+).
Because the marketing plan we did was actually for a startup called “And Chill” (a Netflix recommendation bot), the joke actually plays out as clever rather than as a false way to grab attention. This is the key to sticking the landing on this type of approach.
The subject line never gets enough attention, but it’s critical; if somebody doesn’t open the email, you never even get a chance to impress them with its contents. However if they are ‘tricked’ into opening it, you might lose them forever.
Because it’s so important, you should at this point be looking for evidence of testing. In this case we are — for this specific email we ran an email subject line test:
This result alone should be enough to convince you. If we had gone with ‘[NEW]’ in the subject line we’d be getting a lower open and click rate, making this email drop below the 10% mark.
While we’re at it, we can also take a look at where the clicks came from; what links are people clicking on? As you can see below, a not insignificant amount of clicks come from the second and third blog post in the email.
Now because we’ve added UTM parameters to these links, we can actually track these through in Google Analytics and see how this traffic did.
We can see in the past couple of months this newsletter has been the best traffic driver; over 548 sessions. Interestingly though, it seems that different posts have different levels of engagement from the traffic they drive. For example the event-tracking post only kept people around for 1:52, whereas the paid ads post kept people engaged for over 5 minutes.
So let’s head back to MailChimp and see how well these types of emails are doing in general.
This is great overall – we’re way above (over double) industry averages for open rate, even on our worst performing emails. In some cases we get more clicks than the average email gets opens. This is great to see and a sign that we don’t need to change much here.
However, let’s dig into some of those emails individually, in the campaigns section.
Not bad; anything over 5% CTR is pretty decent for a newsletter in the B2B space, and the fact that we’ve moved the needle above 10% on a couple of occasions is a promising sign.
Let’s dig into those best performing campaigns. The second one down is the ‘Marketing Plan and Chill’ campaign, so let’s look at that 14%’er down at the bottom.
Subject: We Spent $250 on Reddit Ads So You Don’t Have To ?
Great headline; a lot of people are curious about Reddit advertising, and being transparent about how much we spent makes this an interesting enough post to click through on. The thing I think that really swung this was the part of the subject line “…so you don’t have to”. Straight away this implies it was a failed experiment, and it’s human nature to watch a car crash.
But enough self-congratulation; how about the ones that didn’t do well?
Hmmm, ok… although I thought it was a good idea at the time, I can see in retrospect why this didn’t work as well as the others (5.6% CTR). The other posts were helping them do something they need to get done, or letting our readers peek behind the curtain. Meanwhile, this is the equivalent of us telling them to eat their vegetables (and we just happen to sell vegetables!).
This is our worst performing email; only 4.2%. It looks like a similar issue to the first one. We are in sales mode here, pushing tactics for Qualaroo.
Lesson learned: give them what they want and be careful pushing on them what they need.
Let’s get a little bit more advanced here. One really effective tactic we employ often is to set up a sequence of emails that ‘drip’ out to the user one by one based on how it has been since they joined a list. Remember that email from the beginning when we tested the signup flow?
That’s just one of five emails that you’ll get over the course of a week.
Now obviously the first email has a very high click through rate; when someone has just signed up for something, they’re very motivated to follow through.
What makes this interesting is the additional clicks we get from the following four emails. If we’d just left it at one, we’d be missing out on hundreds of clicks.
You can see from the rules here that it’s set up to only send the next email a number of days after sending the last, but only if they didn’t click into that last email.
What you can’t see here is that if a meeting is booked, HubSpot tells MailChimp via Zapier to unsubscribe that person from this list.
This is a pretty cool setup – most businesses forget to do things like this and either don’t send drip emails or send them out to the wrong people at the wrong time, deflating the sales process.
In fact, we have a number of integrations via Zapier which do smart things to automate our sales process. This is pretty rare to see, but it’s something most businesses could benefit from.
Ok, back to the emails. That first one; let’s critique it.
This follows all the best practices; it’s a personal email; it is from a real person; and it has one simple call to action — book a meeting. We include some social proof in here around our Planner tool and the millions of dollars of performance data we have. The call to action is repeated, and there’s a clear unsubscribe link so we don’t get market with spam.
The next email softens the tone a little bit; it repeats the same call to action, but talks more about helping you find what you’re looking for. We haven’t really tested anything different, so it remains to be seen how this works vs turning up the pressure – something to test in future.
The third one is really clever.
It forwards the first email we sent, just adding ‘Any thoughts on the below?’. This technique is more related to how a real human acts, and can break through the normal ‘this is an automated email – ignore’ response drip emails sometimes get. Obviously if they respond they do get through to the real Wilton, so there’s nothing wrong about using this little trick.
If the last one doesn’t get them, this one does. Despite being 4th in the chain, it has one of the highest CTRs at 7%. The thing this does well, I think, is switching up the offer from agency to technology. For a lot of startups, working with an agency is taboo; if they can’t grow their startup ‘without spending a penny on marketing,’ it’s seen as a sign their product isn’t good enough.
Of course this is a load of crap; marketing can’t save a bad product, and a good product will only grow faster with the right marketing. However I think it’s because of this stigma that talking about our technology acts as a real differentiator here.
Often when people start talking to us about our tech, they admit they don’t have a team to execute the tactics our Planner recommends, so there is a natural upsell to our agency. We should consider pushing this earlier in the cycle as a future test.
The fifth one is my favorite; the ‘Break Up’ email.
This is the ‘Hail Mary’ play – at this point the lead is probably lost to us, so we may as well have some fun with it. The use of humor might not appeal to everyone, but then only a small percentage of leads will see it, and maybe this is different enough to cut through.
Now let’s switch over to Intercom and see our behavioral email setup. Unlike our MailChimp emails which are triggered when we choose, or on a specific cadence, these are emails that trigger when the user exhibits a specific behavior.
This type of outreach can be incredibly effective because it’s supremely relevant. Right when a user needs a message, they get it, nudging them towards a desired action. It doesn’t even have to be an email; with Intercom you can send in-app messages and other services can offer text messages or push notifications.
These notifications can also unfortunately be a little more complicated to set up. You need custom tracking of key behaviors and customer segments so you can trigger emails at the right time to the right people.
Let’s see what we have.
Let’s take a look first at the one at the bottom. It has been sent the most times and seems to have a high open and reply rate (54% and 10% respectively).
From this screen we can see what the message looks like; just a simple automated message that shows on the very first visit, for new clients. This is good practice just in general; letting your customers know you’re at their service increases the chances they’ll contact you rather than just quitting when they get stuck.
It also looks like users get an email when they’ve signed up less than 1 day ago.
This email has a remarkable open and click through rate – 60% and 9% respectively. This is a great email to send because it nudges users in the right direction by giving them a list of activities they should be taking on their first day.
There is a clear set of instructions and it walks you through what the product is actually for. The other thing I like about it is there are clear screenshots of what to do, so the user can visualize and familiarize themselves with the platform.
Unfortunately it looks like we haven’t been doing any A/B testing with this email; we could be improving performance here so testing this email will be critical.
One other issue I found, is that none of these buttons have UTM parameters; this means we can’t separate out this traffic in Google Analytics to see what happened after they clicked.
In the same vein, we do have a goal set up, ‘create_test’, but it’s never been triggered.
Knowing that at least some of these users have created a test, this strikes me as a bug, which we should raise with our development team; it’s very possible we’re not tracking this properly.
Related to this, I can see an email that’s called “You’ve launched your first test!” that has never triggered for the same reason.
It’s really common to see things like this, so you should explicitly look for them; lots of hard work went into writing this content and creating this email, and it was all wasted because we made a mistake on our tracking setup.
Now let’s take a look at an email that IS driving a lot of goal actions.
Wow – 29% complete the goal – what was it exactly?
Ah – ok so that’s less impressive. A lot of those people who were sent this email were probably going to log back in anyway; we should probably chose a better goal metric here, like ‘became customer’. The relevancy of metrics is always relative however, for a dormant reactivation this is actually a good goal to have, say for example for this email:
Getting 14% of people who we haven’t seen in 30 days (meaning their free trial came and went 2 weeks ago) to actually log back in, is really great.
However this email is a little dicey; you can see there was a relatively high failure rate here, mostly because a lot of users unsubscribed. From the few users that responded to it, I know this is probably because they weren’t happy with the heart-stopping “You’re fired” subject line.
In this instance we actually added the emoji to help solve this problem and the unsubscribe rate has gone down; this way we get the benefit of the shocking headline, without scaring people into thinking they are actually getting fired.
Let’s jump from the least engaged users to the most engaged; what do we do for users that log in a lot? Our VIPs.
This is a really high open rate and reply rate – we’ve actually gotten some really solid product information from this segment, and some of them have become customers after talking to us.
Crucially, this is a highly targeted niche; they’ve logged in more than 10 times, and are between 7 and 14 days into the product; i.e. in the last week of their free trial.
So what email tactics didn’t work as well as expected?
Here we’re trying to get someone to create their first test, if they’re 4 days into the trial and haven’t done so. This type of tactic can work really well to get someone to experience their first ‘aha’ moment. For us, this is launching a test. If you haven’t done that, you’re not really using our product for what it’s supposed to be for.
Given the importance, only 4% clickthrough rate and 2% reply rate is too low. We haven’t A/B tested this and we really should work that into our plans.
We can obviously keep going for a while here, but I think you get the picture; it’s simply a case of looking at the data, finding something interesting (higher or lower performance than expected) and then digging into the campaign to see what might be causing it.
One more before we finish up.
This is one of my favorites. For a quick in-app message that took 10 seconds to create, we’ve driven an enormous amount of engagement; 47% of people searched for tactics after seeing this, and 11% replied.
We just talked about ‘aha’ moments and this is a small but crucial one; it’s only after you search for tactics that you notice just how useful our tactic database can be. It’s quick wins like this that really make CRM such a cost-effective channel.
And that’s it — a full audit of all our active email campaigns, from our newsletter to our sales prospecting drip and our CRM messages. We learned a lot in this process and uncovered a ton we can test.
Too often, marketers will run email campaigns and leave them on, forgetting the power of A/B testing in this channel. We’re going to be doing a ton of that on our end in our 2017 marketing plan, so stay tuned for more!
Sign up to our newsletter to stay up to date with all the latest movements in the field.
More from Ladder Blog
The art world is being disrupted by generative AI, and artists aren’t happy. Generative AI models like DALL-E, Midjourney, and Stable Diffusion were trained by scraping millions of images from the internet, without permission. Artists claim this violates copyright, while the companies training these models argue this falls under fair use.
Read More →In the world of digital marketing and data-driven decision-making, creative testing is a pivotal tool in achieving business growth. Gone are the days of relying on gut feelings or guesswork; now, business decisions are powered by data-validated insights, meticulously collected, analyzed, and validated. This transformative process empowers businesses of all sizes, from established enterprises to budding startups, to thrive in an ever-evolving digital market. This article looks at the practical applications, challenges, and innovative transformations associated with creative testing, offering you valuable insights and actionable strategies to implement in your own digital marketing efforts for achieving growth and success.
Read More →
