How to avoid running online advertising experiments
January 20, 2015
Let’s say you’re trying to figure out how expensive it will be for you to run ads on, say, Twitter. How do you spend as little as possible while still getting a sense for whether this is a viable channel?
Step 1, is, don’t. Before you spend anything, answer these questions in detail. You can actually get a lot of information about the viability of your plans, without spending any money.
1) Start by estimating or figuring out what the cost per lead / acquisition is for your other channels.
Does it cost $100 for you to get a sale through cold-calling? Or $5 to get a lead through search? Numbers like these give you a number to compare your results to, which is critical for your overall marketing strategy.
And if you already have a cheap channel in place, should you consider doubling-down there instead? What is the scale-up potential of your existing channels?
2) Estimate the lifetime value of one of your customers (CLV).
There are tons of articles out there on calculating this, and there’s even this page which will help. Put this together with the information in (1) to figure out what it’s actually worth for you to acquire someone.
Maybe one of your acquisition channels (like events) is great in terms of volume, but it actually costs way too much money and you could improve results by optimizing that channel first.
Or alternatively, maybe one of your acquisition channels definitely has a positive return, and you could spend more there, first.
You start with (1) and (2), ideally. If you can’t calculate, or at least estimate, (1) and (2), how will you know if your online spend is effective?
But maybe you just want to get a sense of how much online advertising will cost, first, in that case, you can still answer (3) and (4).
3) Figure out what resources you have to devote to online advertising.
Online advertising isn’t “fire and forget”. Someone needs to spend time monitoring results, trying new advertising and variations, optimizing audiences and landing pages, designing ads, and on so on. Do you have someone who can work half- or full-time on this, if your experiment is successful? Or can you afford to hire an agency, and can someone on your team manage that relationship?
If not, you might not want to bother doing the tests to see how effective it is, especially if there are other high-ROI activites you already know about.
4) Figure out the size of your audience.
Part of the appeal of online advertising is that it’s very scalable. It doesn’t take much to reach 10x as many people, and you’re theoretically able to reach everyone on the internet.
But you’re going to be targeting your ads to specific audiences with specific intents. If you product serves a narrow niche, you may find that you actually cannot spend enough money on online advertising to make it worthwhile.
Tools like Google’s keyword planner can help, and most ad services will give you a sense upfront of what the reach of your ads will be. Other channels, like search, can give you clues as to how much people care about your content, too.