6 Ways to Test Your Digital Marketing Strategy
As digital marketers, we all know testing is really important. Testing teaches us so much about our audience, our content, and our operations. The results we glean help us make smarter decisions so we can find faster ways to achieve our goals. We’re pressure-testing our strategy and our assumptions. And honestly, it keeps things fun. Beyond the variety it adds to your day, who can resist the temptation of a surprise success, the opportunity to find something no one else knows, or the next big thing that leads to your career growth?
But how do you actually test your digital marketing strategy? What's your testing approach? Here's a testing strategy that you can apply to all digital channels — social media, website, email automation, you name it.
In general, I spend about 80% of my team's resources (effort, budget, program types) on executions we know will work. We use the remaining 20% of our resources on testing. And by testing, I really mean trying new things. While sometimes it's an A/B test, other times it’s as simple as trying a new ad unit, developing a new audience segment, or running a new type of email.
For example, when Facebook came out with lead forms, it was a surprise. We weren’t planning on them so the point is they weren’t in our plans — and we weren’t counting on the results. But we hadn’t spent all of our Facebook budget yet, so we carved out some money and just tried it with no expectations. Wouldn’t you know, Facebook lead forms become a key component of our demand generation strategy.
This 80/20 may look more like 70/30 or 90/10 to you, depending on your resources. But it helps if you have a general rule of thumb on what you can realistically allocate toward what’s shiny and new.
80% should bring 100% of your results
Testing means you don't know what the outcome will be — and that means your efforts could succeed or fail. In order to make sure you meet your goals and still test, aim to meet 100% of your goals — or close to that — with the 80% I mentioned above. What does that means in practice?
You want a well-oiled machine that helps you meet your digital marketing goals reliably. If your test works, that 20% is just icing and then hey, you become a superstar. If the test doesn't work, no big deal. You learned. And if the test kind've works, the results are incremental — your experiment might not win you any awards, but you emerge better informed.
Give yourself time
In general, we know the bigger the test the longer you need to see results. But I find that you want a good quarter to run any given test. This gives you enough time to set the program, get creative ready, run the program, set up tracking and see results.
Keep in mind, however: you're doing something you haven't tried before, so things can take longer than you planned for initially and/or something unexpected may come up.
Give yourself a budget
If you're testing a paid campaign, give yourself a minimum threshold of what you need to spend on any test to make it worthwhile (aka statistically significant). Use that same minimum threshold for your tests so you can compare them to each other. This amount should fit into your budget — into that 20% that I mentioned above.
You may disappoint a vendor who's trying to get you to spend more upfront but hey, it’s your neck that’s on the line, after all. You're the one with the money, so you should only commit to what you're comfortable with.
Get support on a big bet
This tip almost goes without saying, but if you want to take a big bet and spend more than that 20% in time or budget or both, get your manager's support. This will help because when testing, you need to stay objective. With support from above, you won’t worry needlessly about an experiment’s success or failure. And if it doesn't do well, a manager’s buy-in can help you feel confident when you state that all didn't go according to plan.
There's more than one way to measure "success"
When you're testing, you're evaluating two things. First, does this program help meet my goals? Second, is it operationally feasible? The operations part of the equation is often overlooked. Sometimes we just need to learn how to do it before we can learn what makes it works.
For example, when we launched those Facebook lead forms, we learned how to get the leads to Marketo and Salesforce while we learned what content, creative, and campaigns were best suited to that new format. At the end of the test, all we wanted to answer was: Can we run this program again? Do we want to run it again? Fortunately, in this case, the answer was yes.
At the end of the day, you need to know your levers — the tools you can use to meet various business goals. Testing builds a discipline that reinforces your strategy. You’ll develop a deeper knowledge of your customers, your business, your industry, and ultimately those levers so when you need to meet any given goal, you have options. Happy sleuthing!