Sit down, bad: what is wrong with your autofunnel

Views:1842

Autotronic sales funnel is an automatic promotion of the client with trigger messages from the first visit to the purchase.

In other words, autotunnel independently, without human resources, promotes all visitors to the site to purchase, no matter how many 500 a month or 5 million. With autotunnel, you don't need to hire more employees to work with visitors - everything is done automatically.

So, the studio has seven most common conversion killer mistakes.

Mistake #1: Creating the entire auto-funnel at once

Particularly diligent marketers want to set up the entire autofunnel at once, prescribe all scenarios for each stage of the funnel in as much detail as possible, launch it and never come back to it again. Too bad, but it doesn't work that way: building the entire funnel will take a very long time (with special efforts, this story can take up to half a year), during which you won't interact with customers at all. In addition, all funnels are built based on hypotheses, which can turn out to be wrong. In this case, you'll have to spend even more time fixing the errors, and user interaction will continue to be put off indefinitely.

How to fix it:

Act in small steps: hypothesize, gradually lead users to each targeted action, look at metrics, and repeat this for the next step. Keep the hypotheses that work and refine them, discard the ones that don't. That way you can gradually build out all the messages for each step of the funnel and get small but real results in the early stages.

Mistake #2: Segmentation for the sake of segmentation

Roughly speaking, user segmentation is the division of users by certain criteria, such as gender, age, geography, interests, and so on. Sounds great, but if you don't understand the specific benefit segmentation will bring to customers, it will be meaningless.

How to fix it:

You need segmentation to make valuable offers to customers based on analytics and to increase conversion to targeted actions. Be sure to evaluate the cost-benefit ratio so you don't waste resources.

Mistake #3: Testing on small segments and working with small numbers

Another problem for diligent marketers who want to do the entire autofunnel at once and prescribe detailed separate scripts for each user segment.

The problem is that hypothesis testing on narrow segments will show statistically significant results after a long period of time.

How to fix this:

Start by developing shared values and common messages on large base segments, such as all users. The larger the base, the faster you'll know if the hypothesis works or not. If successful, you can start testing similar hypotheses on narrower segments.

Mistake #4: Ignoring basic math

You can spend a lot of time and effort on an autofunnel that won't produce the results you want, it's painful and not worth the effort and time spent.

How to fix it:

Think ahead and figure out, based on average conversion rates, how much targeted action a funnel can bring in. Is the game worth the effort?

If you want to delve into math statistics and calculate the possible outcome (what sample is needed for the expected outcome and how to properly estimate the hypothesis), you can use Evan Miller's calculator and read articles on the subject.

Mistake #5: Test the buttons, not the values

You can endlessly experiment with the design of auto-messages: change the arrangement of buttons, pictures, and so on down the list. The problem is that such tests may give you a plus in conversions, but they take a lot of time and don't add value to the user.

How to fix this:

In the first stages, focus on experiments with value to the customer: why the customer buys your product, what tasks it solves, what's more important to them, because such experiments yield multiple changes. That way you'll figure out which values work best faster, and continue to experiment with narrower things like picture design.

Mistake #6: Wasting developer resources

In an effort to make everything perfect, diligent marketers bring in other specialists to help with integration, layout, design, etc. This is not the smartest approach: the hypothesis may not work, and developers' time and efforts will be wasted.

Developers are the most expensive resource. Always try to use your own resources during the hypothesis testing phase.

How to fix it:

Turn off your inner perfectionist and run the hypothesis in a minimum viable version format. If it shows good results, get the developers involved and make it look good.

Mistake #7: reinventing the wheel

Some marketers tend to create something unique and outstanding instead of making a workable version using the scripts and tools that will bring real results.

How to fix this:

Pay attention to the experiences of people who have already solved similar problems.

Steal success stories, see how your competitors and partners are doing, how companies completely different from yours are doing. Sometimes the most trivial things get the best results.

Was this page helpful?