 |
Start Cases/Testimonials Why us Tariffs/Guarantees Implementation Pro et Contra Contacts |
|
+81%
+67% 
+51% 
+68% 
|
Our cases: conversion rate growth of 1.39 ~ 1.81 times (by 81%)
Growth guarantee: up to 1.61 times (by 61%) Cost once-only: from 450 USD Result: from 4 weeks |
|
Try it now! We'll call and tell the details. |
|
You are
 online store, LP
 Internet service, SaaS
 content website, forum
 any commercial website
|
We are
 conversion rate growth service
using:
|
You gain:
 conversion rate growth up to 2 times
 growth guarantee up to 1.61 times
 fast test completion
 payment based on results!
|
|
we are not an A/B testing
We don't do simple A/B testing or classic MVT Our methods much more powerful than A/B testing or MVT because we are using millions of page variants and can handle large number of combinations |
we are not an outsourced analytics
We don't work with historic data, accumulated without us and given «for analysis». Live interactions with your site's visitors is the key component to our success |
we sell not an «experience», but testing achievements
Testing methods - is our strong suit. We have an experience, too. But it differs from those of present «LP improvers». They seek for parameters of strong effect, we successfully use any of them. |
Growth of 1.51 times (+51%)
Target: conversion into appications
4.33% → 6.55%
Pavel Zhavnerov, pzhav.ru case description of pzhav.ru (have been finished too early) |
Growth of 1.81 times (+81%)
Target: conversion into click (advertising)
1.93% → 3.49%
Eugene Letov, poiskum.ru case description of poiskum.ru |
Growth of 1.67 times (+67%)
Target: conversion into click (advertising)
0.52% → 0.87%
Serge Baryshnikov, bigpicture.ru case description of bigpicture.ru |
Growth of 1.68 times (+68%)
Target: conversion into click (advertising)
1.61% → 2.71%
Serge Panchenko, muzhzdorov.ru case description of muzhzdorov.ru |
Growth of 1.24 times (+24%)
Target: conversion into SaaS registration
9.88% → 12.22% 
Roman Tsapin, insales.ru case description of insales.ru (have been finished too early) |
Growth of 1.39 times (+39%)
Target: conversion into click (advertising)
0.32% → 0.45% 
Aleksey Dobrodomov, buhgalteria.ru case description of buhgalteria.ru |
Growth of 1.52 times (+52%)
Target: conversion into click (advertising)
2.38% → 3.6% 
Eugene Letov, chelium.ru case description of chelium.ru |
I planned for a long time to make an AB-test to increase income... For some months KONVR automatically tested huge number of advertizing designs. As a result - i got a design of advertizing blocks, raised number of clicks of 2 times by eye, at the same traffic. Using this system anyone can test all the things which have any influence at conversion rate. For advertising test starting setup took about 1 day. I recommend KONVR! |
... we turned to ad agency "Papa Carlo" for improving of conversion rate of a main page www.insales.ru ( SaaS online store platform ) After multivariate test with a KONVR technology we worked with 15000 landing page variations, as a result we got a best variant and improved CR by 25%. We thank "Papa Carlo" employees and Eugene Trofimenko (konvr.ru) for a good job! letter PNG |
Pavel Zhavnerov
pzhav.ru
chartered clinical psychologist, PhD I can say that the developed system - is a best thing ... it is able to detect the best variant, even within a testing process I started to understand a selecting process and nearly understand why different changes have an effect. ... System is useful for any business and I don't say maybe, I say yes, this approach always gives good results, conversion rate raised week by week, it's the key to successful system, thanks again for the result! |
Serge Panchenko
muzhzdorov.ru
I met Konvr founder at a KINZA-2014 conference... I turned to him with an order to optimize the ad units of an on-site advertizing (Yandex) ... Prior to that site with good attendance gave insufficient in my view revenue. New optimized advertizing blocks showed higher conversion results compared with my settings. The conversion rate increased by about 50-70%, which has increased my income as well. I'm very pleased with cooperation with Konvr ... sure I will refer to this system the future and I will recommend it. |
| free | startup | business | enterprise | agency | |
| Payment plan, USD | Choose it |
Choose it |
Choose it |
Choose it |
call us |
| Cost by results of each 1% conversion rate growth, USD | 0 | each 1% CR growth<30 USD | each 1% CR growth<60 USD | each 1% CR growth<120 USD | depends on volume of sales |
| Conversion rate growth guarantee |
no guarantee | growth guarantee by 15% (of 1.15 times) | growth guarantee by 31% (of 1.31 times) | growth guarantee by 61% (of 1.61 times) | no guarantee |
| Expected conversion rate growth | depends on you | expected growth by 15% ~ 30% of 1.15 ~ 1.3 times | expected growth by 31% ~ 60% of 1.31 ~ 1.6 times | expected growth by 61% ~ 100% of 1.61 ~ 2 times | depends on you |
| Continue the test when growth is over expected limit | - | no | no | yes, continue for best result | agreement |
| Website pages number | one landing | entire website | entire website | entire website | from 5 websites |
| Payment by results | - | yes | yes | yes | agreement |
| Testing parameters design and implementation | your own | yes | yes | yes | by your specialists |
| Expected test time | less than 2 weeks | 4 weeks | 4-6 weeks | 4-8 weeks | up to 8 weeks |
| Changeable page elements number | up to 4 | up to 8 | up to 16 | up to 32 | up to 32 |
| Unique page variants number | up to 256 | up to 6561 | up to 43046721 | unlimited | unlimited |
| Computing time units | 1 | 4 | 16 | 64, dedicated server if needed | agreement |
| Support | - | e-mail, skype | e-mail, skype, phone | agreement | |
| Tariff elevation is allowed when testing | no | no | yes | yes | agreement |
| Daily monitoring and correction by specialist | no | no | no | yes | agreement |
| Gifts | no | not today | gift: test "startup" |
gift: test "business" and usability report |
|
| Prepayment, USD | - | 450 USD | 450 USD | 990 USD | agreement |
free |
startup |
business |
enterprise |
call us |
1 week
|
2 week
|
3-6 weeks
|
result -
|
Others |
We |
|
«Others» need to find page elements that greatly affect the conversion rate.
... And what if there are none of them? Their approach is simple - to find a factor having strong effect on the conversion. Better if it immediately fold increase in conversion rate. But most page parameters affect the conversion weakly, will they succeed? No way. If page have been optimized before - it is likely to have strong parameters previously found.
|
The right approach - to summarize the effects of «weak» elements
... They are guaranteed to exist! First, their existance nearly guaranteed. Any of page elements affects on conversion rate. Absolutely.
But in different ways - some have strong effect, others weak.
The only problem - the weak factors are difficult to work with. We solve this problem using interactive technologies. |
|
«Others» can't work with really large number of page elements in test.
They use the classic static approach - each unique combination of elements values must accumulate significant statistics (usually about 100 actions). Otherwise - you can get almost random result. Your traffic is not infinite, and is not enough to accumulate such a statistics for each combination. So: You have NO freedom in the choice of dozens of page elements for the experiment. You must limit yourself in a number of parameters. And, of course, You won't get f strong growth of conversion rate, too. |
We are able to work with large number of parameters using interactive technologies.
Interactivity, which is the basis of the method - when we determine by ourselves - at each entry of a website visitor - which set of element values will be shown to the visitor at this moment. The action of this impression (or action absence) immediately affect the appearance of further hypotheses. We do not work with displays that are accumulated by someone else. We manage the impressions and try to show promising options to accumulate good statistics on hypotheses more often and more quickly. There is a similar method of «multi-armed bandit», but it can not really work with a large number of parameters and unique combinations. We successfully solve these problems and showing best cases of conversion rate growth. |
Other services |
Konvr service |
| question, objection | answer |
| Millions of pages! This is ridiculous. Creating of millions Landing Pages - it's a big work, no one will do that. | Millions not separate LP's, but millions of unique combinations of variable elements of the page.
All of them are created automatically, not manually. MVT (Multivariate testing) is simple - there are 10 header variants, 10 button variants, 10 different pictures - eventually have 10 * 10 * 10 = 1000 unique combinations. Yes, they are lot of them, but they are very similar to each other. |
| But for reliable testing of millions of pages is not enough traffic, no matter what you promised! | Wait a minute! We are not saying that we reliably check all those millions. Reliably tested only the original page at the beginning of the test and the best page - test result when finished (they are tested for a week, during which time they both need to accumulate more than 100 conversions).
All other combinations of parameters - a complex cocktail that is used by machine learning. |
| But with millions of pages and little traffic may even occur that many pages do not get a single impression! How to use them for learning? | First, we are able to work correctly with small statistics and get useful data from it.
Second, not all the combinations we want to show. Combinations of element values that are promising and interesting to algorithm, surely gain their impressions and actions. Third, information on the interrelation between the parameter values we get anyway. |
| Well, I understand. Surely interrelations don't exist at all, parameter values are independent. We can make a lot of shows, selecting parameter values randomly, calculate the conversion rate for individual parameter value for slices - and choose the best values. Without your service, yo-ho! | Even if our parameter values are independent, you get almost a random result. When we analyze the parameter values separately from one another - you will take into consideration that the "tail" of the other parameters makes a very large random contribution into result (there are dozens of parameters). It is much greater than the contribution of the analyzed parameter. This is described in an article with examples and numerical calculations. Still, you need to analyze a lot of website visitor's actions, which you do not.
Second, the interrelation between the parameter values - is a reality and can dramatically increase conversion rate. |
| OK. But still there are a lot of pages! You write about interactivity, but also "multi-armed bandit" does not support such a number of pages! | Yes, our technology is more sophisticated and more powerful than a common multi-armed bandit, in which the number of "handles" the order of 10 ~ 100. We support much more variants when working interactively. |
| How can you ensure an increase in conversion? Nobody does guarantee the result of work. | Others don't guarantee, because they work with a small number of strongly influencing parameters (that need to be found first). What if strong influence doesn't exist?
But we work with a large number of weakly influencing page elements, but they are guaranteed to exist. By the way, so with the increase in guarantees is increasing and the number of parameters used (and calculations complexity, and cost). The only question is technical issue - it is more complex to identify good values, we successfully do it. |
| Anyway, I do not believe you. Unrealizable promises ... | Why, it is easy to check, and you can do it yourself. We describe experiments on the user model and examples of solving a model. You can create your own robot and your own model. Robotic traffic is free (but in this case the experiment proceeds without expert participation). |
|