We take full responsibility with taking payment for the increase in conversions only
Cost per click is rising, but your conversion rate is barely 1% on average? We take responsibility for the conversion rate on the site. We will ensure the growth of applications with the same advertising budget. You pay only for conversion growth.
Split testing (A/B testing) - testing different versions of pages on the of the site and their impact on sales.
Versions may differ in the layout of elements, content, forms, CTAs.
Our job is to find the elements that affect conversion rates and improve them. We will create a test variant ourselves and test its effectiveness. All our hypotheses are based on data from analytics systems and research into the specific behaviour of users on your site.
Once successfully tested and implemented, you'll see an increase in conversions and revenue on an ongoing basis. And you only pay for it once.
Efficiency is at the heart of our programmatic advertising service, featuring strategic planning, real-time optimizations, and ongoing tracking of key performance indicators (KPIs). By fine-tuning bids and enhancing ad creatives, our methodology delivers streamlined programmatic campaigns designed for peak performance and impactful engagement.
On average, between 6 and 10 split tests can be performed per year.
In a pay-for-growth tariff, we do not agree on split-test ideas, as we have a direct financial interest in conducting only successful tests. If our hypothesis does not work, we simply earn nothing.
However, there are cases when the client wants to participate in the selection of ideas for testing or to test their ideas. We have a separate tariff for such a case.
The split testing methodology assumes that half of the audience will see the site as it is now and the other half will see it with changes. With the help of a special system, we conduct testing and the system tracks which version of the site has more targeted actions. If the version with the changes we came up with and implemented wins, the agency gets a reward. Its size depends on how much the conversion rate has increased. If our version did not increase this indicator, the agency does not receive a reward. Split testing is a search for solutions to increase the probability of an order from the current audience. We don't know in advance if a hypothesis will be successful or not. But usually 4 out of 5 hypotheses are successful. *If the desired target is not set up, we will set it up and start split testing immediately, as the technology allows us not to wait for statistics to be collected.
Three main sources:
All elements, including text, forms and prices. At the same time we stay within the design standards of the site. The more restrictions and prohibitions on changes, the more difficult it is for us to analyse and generate ideas and, consequently, the more expensive the point of increase. We agree this before we start work.
We do not coordinate, but we inform. The essence of the service is that there are no delays in the approval stages - we need to test and implement successful hypotheses faster. In the briefing and in the contract we specify what cannot be changed, so everything else can be changed, so there is no point in synchronising.
The cost of a point of gain is calculated on a project by project basis. The speed of the test depends on the number of conversions per day and the size of the change. The more conversions, the faster the tests run and the more hypotheses can be tested in a month. For example, for a project with 200 conversions per month, we will run one split test per month. This is the minimum threshold for projects we are willing to take on. The ideal case is 500 or more conversions per month.
This is why we make a forecast based on the current conversion rate and calculate the payment based on the actual result. This approach allows us to fairly assess the contribution and result of the agency's work. Here's an example. Let's say the average conversion rate on the site is currently 1%. We want to increase it to 1.3%. According to the formula, this is an increase of 30 points: (1,3%/1%-1)*100%, During the experiment, the traffic will be split evenly between the original version of the site and the variation - if 1000 people visit the site during the day, about 500 will see the original version of the site and 500 will see the variation. This keeps the experiment clean: if 500 visitors who came to the original version showed a conversion rate of 1.1%, we count the actual increase from that level. We don't influence the initial version, so the increase is due to some external cause. However, we do influence the variation, and if in the same conditions 500 visitors show not 1.1% but 1.2% variation - it is fair to assume that this is the result of the test. Consequently, the coefficient in the test has actually increased from 1.1% to 1.2% and the reward is considered to be (1.2%/1.1%-1)*100%= 9 points of increase * N rubles per 1 point of increase.
If your advertising campaign is not paying off, you want to supervise a contractor or a specialist in the company, to improve current performance - come to us for an audit. We conduct a complete and thorough audit
I understand that Performiq will securely hold my data in accordance with their privacy policy.
Restrict display locations next to sensitive content if it doesn't align with your brand's policies or values
More infoProtection from real-time scrolling and click-baiting. Screens out robots and you only pay for contact with real users
More infoWe will count the display if at least 50% of the clip is in the line of sight of the user for at least 1 second for a static banner (2 seconds for a video).
More info
