You’ve learned about marketing automation, you’ve created a marketing plan, and you’ve implemented it. What’s next?
You need to know how to review your marketing automation data. Periodically reviewing the data and results from your marketing helps you:
Sounds great, right? Let’s get started:
Figuring out when and how often to evaluate your marketing campaigns is something of a balancing act. On the one hand, you want to review your marketing results often enough that you can identify trends and adjust your strategy before wasting time or money. On the other hand, you don’t want to be reviewing so often that there isn’t enough time for a pattern to form.
Whenever you’re looking at the aggregate data (which is just a fancy way of saying “lots of information”), you need to gather enough data points to reach statistical validity. If you work for an enterprise company that gets thousands of customer interactions every week, then reviewing your data once a month would likely be fine. Looking at a data set with tens of thousands of points can help prevent sampling errors.
On the other hand, if you work for a smaller company that only gets a handful of interactions every week, you’ll have to wait longer. With smaller data sets, one or two interactions ending in a sale doesn’t necessarily indicate a larger pattern.
In general, it’s a good rule of thumb to review your marketing results when:
No matter how often you do a review, your review should consist of:
Your return on investment (ROI) is the amount of money you make after spending money on something. It’s usually expressed as a percentage.
For example, if you spend $500 on ads and get $3,000 in sales as a direct result of ads, your ROI is:
($3000 – $500) / $3000 = .83, or 83%
Getting an estimated ROI is relatively easy, as you can see – all you need is your marketing spend and your marketing results. You can read more about calculating your ROI here.
Depending on how tight your resources are, you may want to account for labor costs when calculating ROI. If you’re running an agency or a similar business, where clients are billed based on your employees’ hours, this isn’t really useful. However, if you’re not billing based on hours, then factoring in for the cost of labor when calculating ROI can give you important insights. Something that’s free but very time-consuming may wind up being more expensive than something that costs money, but runs on autopilot while doing so. After all, you are paying your employees for their time!
In addition to looking at your overall ROI, it can be useful to track ROI by campaign or channel (or both).
Campaign-based ROI is great to keep an eye on if you have multiple campaigns running at the same time – maybe you’re running a promotion on your golf-related products, but also running several event sponsorships. Tracking the ROI of those individual campaigns lets you know which was the most profitable, so that you can do more similar campaigns in the future.
Channel-based ROI is useful for the same reason – knowing which channels perform the best can help you create a more effective strategy.
KPIs are “key performance indicators” or specific metrics of success. The marketing KPIs for your business will depend on your goals and priorities. For example, if your goal is to increase customers by 20% in Q3, your KPI would be the number of online customers. If you’re working to improve marketing efficiency, hours spent on marketing per week might be a good matching KPI. At any given point, you’ll likely be tracking between three and ten KPIs that are tailored to your current goals and activities.
In addition to tracking those KPIs, it’s also a good idea to track other, broader metrics, regardless of your current goals. This gives you a benchmark and helps you see overall growth or decline.
While most channels can be evaluated using the common metrics above, as you dive in deeper, you may want to look at some channel-specific metrics as well.
Two examples of channels that have their own metrics are email and PPC ads:
For other forms of advertising or direct mail, most people track success by giving customers coming through those channels a specific URL to go to, a specific discount code to use, or both. You can calculate how well the ads performed by looking at how the visitors to that landing page behaved or how many people used the discount code.
You can also look at the broader metrics like conversion rate and cost per lead, and segment them by channel. Do people from certain channels buy more? Where are the most loyal customers coming from? What channel draws in customers with the highest LTV? By keeping track of these metrics, you can focus your efforts on the channels that get the best results.
After you’ve collected all the data, you can chart it and look for trends. What you’ll be looking at here will likely be specific metrics tracked over time, to see if they’re increasing, decreasing, or holding steady. Some of the metrics you’ll want to look at this way include:
For all of these, you can look at the total number, as well as the metric segmented by campaign or by source (traffic that’s related to your spring sale campaign, for example, or traffic that comes from Facebook).
When you’re comparing your statistics, you may notice correlations. For example, you might notice that a spike in traffic from Facebook coincides with a spike in sales. It’s tempting to immediately assume those two things are related, but if you’ve ever taken a statistics class, you’ll probably remember that correlation doesn’t equal causation. (If you want a visual example – or several – of this phenomenon, check out these charts of spurious correlations.)
Once you’ve collected all of this data and compared it to previous marketing audits, you can review your assumptions and theories to see how they line up with the data.
For example, let’s say you created a customer persona. Based on that persona (which was hopefully created with data, but if you’re a newer business or don’t have the resources to do market research, you may have had to make a few guesses), you assume a lot of your potential customers are hanging out on Facebook. You suspect that these users will convert more readily than generic SEO traffic.
To test this theory, you create a marketing campaign that uses organic traffic from Facebook and Facebook shares, as well as paid Facebook ads. After three months of running these Facebook campaigns, you may review the data and discover that your hypothesis was correctL the traffic from Facebook converts at a higher rate than organic SEO traffic. Building on that, you can begin planning other ways to test and optimise your Facebook strategy to drive traffic and sales.
But maybe you discover that the traffic from Facebook doesn’t convert as well as organic SEO traffic. In that case, don’t think of this experiment as a failure – instead, think of it as a learning opportunity. This is a great time to think about why you came to the assumption you did. With this example, you might ask yourself questions like:
Maybe you need to try a more organic approach, where you join Facebook groups and focus on being genuinely helpful (occasionally mentioning your products, without too much self-promotion). Or maybe you can take some of the resources you had previously allocated to Facebook, and use them to test a paid Twitter or LinkedIn strategy.
The great thing about marketing automation is that, over time, you’ll always collect more data. By looking at the data you have available, creating a theory based on that data, and then testing the theory, you’re constantly iterating towards a more successful marketing strategy.
Now that you’ve looked at all of the data and at your previous assumptions or theories, it’s time to look at your goals. For your previous goals, look at:
While setting goals for the next month, quarter, or year, ask yourself:
Reviewing not your marketing results and your goals on a regular basis ensures that you’re consistently moving the needle on the metrics that matter. After all, that’s what marketing is meant to do.