Marketing decisions can cause a domino effect, especially the changes directly affecting your site. That’s why bigger decisions often end up in committee, or someone up the ladder chimes in about a headline to change, a section to delete, or a video to include.
It doesn’t help matters when an already tough decision has a hippo pushing back. That’s not the river-loving, thick-skinned mammal, but it’s seemingly just as tough to fight against. A “HiPPO” here is an acronym: Highest Paid Person’s Opinion.
Far too often, the person a few rungs up the ladder says something like, “we should change X because it is not working.” She might be right. But what if she’s wrong? How can you confidently speak up?
This happens across industries. It happens in small teams and in large companies. It’s nearly universally difficult to push back when someone “bigger” has spoken out.
You push back anyway. Until there is proven data, everyone is on the same level–just guessing.
Turn your approach from “We think we should do X” to “Here’s what the data says we should do next.”
It’s hard for traditional marketers to give in to pure data-as-a-decision-maker because their gut has served them well. You might be one of them. I am one of them. That’s why I will throw in a caveat that the data doesn’t stand alone. Your business success relies on having talent to write compelling copy, conceive art direction, and come up with the ideas that connect with your audience. The data confirms (or redirects) how you use these talents.
Your data is your best adviser. It shows you what is and isn’t working and where your profits are coming from. It gives you the insight you need to give your customers a worthwhile and productive experience on your site. If you do that, your customers will stay longer (or come back), they’ll tell more people about you, and they’ll even write about how great your products and services are.
How to Get Data
- Test and measure what’s there
- Research what’s been there
Measure What’s There
The first case involves active testing of the choices you make. Test elements of your sales funnel–from ad copy to page count to pricing, test emails and the number that you send, and test ads against each other. That’s researching “what’s there.”
Here are some ways to get started.
Google Analytics — It’s pretty much the starting point for all website testing. Most site owners start with Google Analytics to discover visitor engagement and conversions. It’s thorough and gets into the gritty details, but that rabbit-hole type navigation can grow unwieldy. Proceed with caution.
VWO (Visual Website Optimizer) — We use this extensively on SpyFu.com. It lets you create and A/B test different versions of a page on your site to find the version that converts best. It offers multivariate testing software, heatmaps, usability testing, etc.
Optimizely —This tool offers A/B testing and multivariate testing on different platforms, so you can measure performance of your website or mobile app.
Unbounce — Set up new landing pages variations to test, and then test against the winning one. Not everyone has the ability to consistently churn out multiple landing pages, so this tool makes it more accessible.
Research What’s Been There
Before spending your marketing dollars you’ve always looked at reach, impressions, ratings, and traditional conversions. That’s a start, but you can get even more detailed. Research how others before you have spent their ad dollars and what they ran. Find out how long they invested in one particular direction. That information alone tells you whether they considered it worthwhile.
Here are some ways to get started.
SpyFu — Among many pieces of competitor habits you can see, SpyFu lets you research keywords in a niche to see who has advertised on it (and the copy they used) so that you can prioritize your budget on more effective keywords.
Reading into advertiser patterns (buying the same keyword and paying top dollar for it month after month) helps you reduce the risk of investing in dead-end keywords.
WhatRunsWhere — Similar to SpyFu, WhatRunsWhere shows you the data to help make decisions about where to spend your display ad dollars.
AdBeat — Also a display ad research tool, AdBeat helps you research your competitor’s ad placement and landing pages from those ads.
It Won’t Be Perfect
Remember this is part of a cycle. Waiting for “the right data” can turn into a bad habit if it holds up your actions and actual progress. This is where you dig into tools to help you prepare. This is also the time to embrace that talent that you have in your team. Combine all of those insights, make a well-informed decision, and push it into a test. When clear findings come back, adjust and test something new. You should always be testing in a drive toward constant improvement.
We’ve often adopted the idea of being “better than yesterday” to get us past a rut of aiming for perfection. Sometimes you reach a point where you have to say “let’s make it good and test it for better.”
Don’t Kill the Loser
Conclusive data from a test doesn’t have to be the death knell of a marketing decision. Yes, you should test everything that you can, but it’s possible that you’re testing two good paths against each other.
Let’s say that the data tells you to run a short 20-minute video on your landing page because it won out over the 10-minute demonstration. Great finding! Still, that long 10-minute video might be valuable elsewhere. Don’t chuck it. Test it against something else.
Ad platforms are a great illustration of using data to make smart decisions that directly affect your ROI. Focus your ad dollars with the winner; that’s a no-brainer. The nuance is in the second part. Maybe you get a better value when you advertise in AdWords vs Facebook ads, but don’t write off Facebook entirely. Just shift the ratio of how much you’re spending on each one. Reaching that different audience might have a secondary effect on your overall conversions.
Keep Gathering Data
Just because it worked once doesn’t mean you’re locked in. Audiences change. Times change. Pull out an old lesson and challenge yourself again. Can you make it better? Especially if you know why something failed before, iterate on it and turn that into a new creation.
One final tip:
Test your guesses with behave.org. It’s like a gallery of Test A vs Test B experiments and outcomes. You can guess which one did best and see what they learned. Use it for inspiration of things to test and to follow a path of why something worked.
Start backing up your decisions with cold, hard, data. You will move from “I think we should” to “we know what to do.”