How I spent the summer or the battle for conversion on one software site

^{Shot from the film "300 Spartans"}

It is impossible to control that which cannot be measured. Any business is decomposed into indicators. And the software business is no exception. How many parameters affect our profit? How often should they be measured? How to effectively influence them? How to measure response?

At the ISDEF autumn conference, we remember the report of Indigo Byte Systems Director Denis Zhuravlev about how he experimented with his website all summer and how it ultimately influenced the conversion. And now we are happy to share the article-squeeze from this report - go under the cat.

')
Indigo Byte is developing a niche desktop B2B product - Dr.Explain . This is a program for creating help files, user manuals and online help. The product is technologically complex and specific. The main selling point is the site. The company offers the user a one-time license, no subscription. For review there is a free version. Its peculiarity is unlimited use, however, watermarks appear on the result. The average bill - from $ 500. In this regard, the purchase of a product is always a deliberate, balanced decision. Sales funnel is quite long.

Want to manage - must measure

The company's problem is not original: you want more money, money comes from sales, so our goal is to increase sales. Without an understanding of the essence of the process, of cause and effect relationships, it is impossible to manage sales and, therefore, to get more profit. Therefore, a bit of theory.

Converting site

Let's start with finding the answer to the question “What determines the conversion on the site.” Captain Evidence asserts that it depends on many factors, namely:

1. Relevance to the source of traffic and the problem. The information in the advertisement should correspond to the content of the site and - most importantly - solve the client's problem.
2. Clarity and visibility descriptor. The average time to visit one page of the site is a few seconds. Therefore, the page should always have a bright and clear descriptor that will give an answer to the question of how much the user will receive from the site. The scheme of a good descriptor looks like this: This is the _______, for which _________ by __________ (fill in the blanks with the appropriate description for your product).
3. Key benefits and advantages. Zadte pain and the problem of a person describing their benefits and advantages. Remember, a person comes to the site with a specific problem - offer him a solution. Think about customer fears; what he fears, in what he feels insecurity. The information on the page of the site should rid him of them, show the benefits of the transaction.
4. Special offers This option does not always work, it is more relevant for the product in the entertainment sector, where spontaneous purchases operate.
5. Social proof (others use it). Your task is to show that people use your product. Such information helps to make a purchasing decision.
6. Visualization of a product or result. We remember that the user spends very little time on the site, so he is unlikely to read large texts. Therefore, visualization - video, screenshots, pictures - must have. The only caveat - do not use pictures with happy girls from photo stocks, for one simple reason: there are a lot of such photos, and they are fed up with order.
7. Credibility (seriousness, image). This item is very relevant for B2B business. The attributes of credibility include contact information of the company (address, phone, e-mail), reviews about the company. Even if you just increase the font size that the company's phone number is written to, this will increase the conversion.
8. Design. There is no one-stop advice for this item. Experiment and test colors, layout, fonts.
9. Clear Call To Action. This is the most important point. The person who came to the site should be clear what action he should do next; do not spray his attention.

Life hacking from Denis Zhuravlev: "Learn to talk about your product in one sentence, while answering the questions for whom the product is and what problems it solves"

A / V tests or how to make mistakes quickly

Not all experiments end well, i.e. profit, so it’s important to be able to “make mistakes quickly.” A / B test is just a cool way to test your hypothesis quickly and without significant loss in case of an error.

The method is as follows. The traffic is evenly divided in half, and each part of the traffic gets its own version of the page / site. For each version conversion is considered separately. After counting, the result is compared and the “winner” is determined. A / B testing can be carried out a great number of times, as long as your imagination is enough for test options.

Pros:

• There are no sharp failures when testing a bad option;
• Minimal dependence on seasonality. Testing two hypotheses under the same conditions increases the reliability of the study.

Only one minus: good traffic is desirable to ensure statistical certainty.

Life hacking from Denis Zhuravlev: "Read and consider everything!"

Mechanics A / V testing

There are many specialized services for analyzing the results of A / B testing, the most famous of which are Google Analytics. But in life there is always a place for exploits, so Denis Zhuravlev offers his own version.

At the beginning of the test, an experiment table is created. Each visitor "at the entrance" is put a cookie labeled A or B for each experiment. When performing the target action (load) - is registered in the database. After that, it is possible to compare the number of targeted actions with traffic. The screenshot below shows how it all looks in practice. A page containing one review (Option A) and many reviews (Option B) were compared. The first column shows the number of visitors, the second shows the number of downloads, and the third shows the percentage of downloads from the total number of visitors. It is clearly seen that with a slight margin wins option B.

Relevance Testing

We tested two types of ads. Option A looked like this:



And option B is thus:



According to the test results, Option B won, it was he who came closest to the user's fear, providing a solution to the problem of creating a certificate for a program without special knowledge.

Life hacking from Denis Zhuravlev: “Look for real user pain and fear. Create separate landing pages for each of them. ”

Descriptor Visibility Testing

When testing the decryptor, option A looked like this:



View of option B is shown below. The difference was in the size of the fonts.



On average, an increase in font size in headlines led to an increase in conversion from 2% to 35% on different pages.

Life hacking by Denis Zhuravlev: “Increase the font size of the key headings. Even if it breaks the design. Forget about 10 and 12 pins, 30 pt is our choice. ”

Testing Social proof

Your task is to show life around your product. For this purpose, suitable social network plug-ins, video reviews, text reviews, but always with photos. Nuance - the review must be real. Therefore, "do not solicit" the texts of users. Sometimes photos on the background of the carpet only emphasize the reality of the response to the product (this works especially well in the entertainment sector).

The addition of the facebook social networking site plugin to the page showed an increase in conversion of up to 376%. The exception was made only by our French sites, on them there was no such prompt increase in indicators.

Life hacking from Denis Zhuravlev: "In any way show that other people use the product."

Product Imaging Testing

In any way possible visualize your product, post screenshots, shoot video. This is especially true if you are dealing with a complex technological product. In this case, either show the product’s work step by step, or show the result. So that the client's eyes lit up, and he said: "I also want it so!".

In testing, a page was compared that offered to watch a video on YouTube dedicated to the product (option A) and a page without this possibility (option B). In option A, the conversion increased to 136%.

Credibility Testing

Find the differences between option A:



And option B:



In the case of the option In the absence of icons showing compatibility with the OS, the conversion increased by 32%.

Testing Call To Action - One Action

Decide what is the most important action on your site. It is not always a call to buy. In the case of a long incubation period, the noticeable button “Download demo version” is more convincing and necessary.

On the Dr.Explain website, in the case of option A, the page contained two appeals: download the demo version and buy the full version of the program. Option B contained only the call to download the demo version. In this case, the conversion increased by 26%.

Life hacking from Denis Zhuravlev: "Leave on the site a call for only one action - the most important and important for you."

In addition, do not take away the client from the page, remove all unnecessary links, leave only the most important thing - that helps the user to make the final decision.

In the call-to-action matter, even the color of the button matters. In option A, the button was green; in option B, it was red. Conversion in the case of option A is higher by 31%.



Also important is not only the color of the button, but the text on it. The meaning of the word Download is understandable to everyone, in contrast to the word Install, thanks to which the conversion increased by 42%.

Design testing

With the design you need to experiment, experiment and experiment again. In some cases, modesty wins, and sometimes it is necessary to change the location of the blocks on the page.

In the struggle for the conversion of small things there. The more you try and experiment, the greater the chance of success. The easiest and fastest way to get customers is to buy traffic to the site. But before buying it is imperative to check and optimize the conversion. Remember that conversion is important, but, as is the case with any other metric, it is not something absolute, and there is no limit to perfection.

PS You can watch the video of the report here , and buy tickets for the upcoming software meeting for the September ISDEF - here .