As we do Convead: from concept to implementation

we were as ~~drinking~~, discuss issues of eCommerce and complained to each other that all existing Analytics services or do not know something very important or too difficult to configure, and often have to use a combination of several services to achieve their goals. But it is we the people, technically literate, and how many people suffers from the fact that I can't even figure out how to pass event to Google Analytics?

So we came up with the idea to make a service that would help online stores to increase conversion the most simple manner, without requiring any special alterations of the store. And the main thing – that the service itself was very simple, with minimal set of the most important functions.

And what is most important to the store owner? Of course a good conversion! Low percentage of conversion is a major headache for the vast majority of online stores. For this figure I marketers and UX designers, his track store owners as the key efficacy parameter.

Of course, there is the problem of low attendance, but it, unlike conversion, we would not be able to influence. Yes, and the degree of its importance, in the end, still depends heavily on conversions. Imagine that your small Internet shop is visited by 100 people a day (which is very little), but the conversion you have 30% (dream on, yeah) that gives you 33 of the order of the day! Not bad for a small shop with a meager attendance?

In General, reasoning thus, we decided to fight for the conversion! Thus was born the idea of service Convead (the abbreviation of the English. "Convert your Lead"). Our main idea was to create a service that would allow the online store:

the

to know your visitors and to remember them;
properly segment them for any signs;
to see how the online store where it bottlenecks, where visitors fall off (both in General and individual sectors);
to influence these segments in one way or another, encouraging them to make a purchase and increasing the conversion rate;
to evaluate the effectiveness of these actions, modify them if necessary — we call this feature “Analytics And Actions Combined”.

And yet, we have identified the following consumer benefits, which should have and to which we decided to strive for:

the

have a minimum barrier to entry, the most simple configuration, requiring no special knowledge;
require the least amount of time in the day to have maximum automation of effects and intuitive visualization of key performance indicators.

So, we decided to focus on the advanced marketers who know how to set up Google Analytics with your eyes closed, not programmers, can customize your store the way they want, but on a normal average owner of a small shop, let's call him Bob.

Vasya no time to delve into complicated Analytics tools, it is afraid of the word "cohort analysis" and "A/B testing". He wants to simply and clearly explained: how many people came, how many of them fell off, at which stage of the life cycle it occurred, and what the conversion is today. And in an ideal – still and click the magic button to "all was well".

I purposely did not want to go into a description of the competitive advantages of our service over other or compare their functionality. First, we are still in beta, and many features still to do, and secondly the article is not about that. In this and following articles I would like just to describe all the difficulties we faced during the development, we have taken technical and organizational solutions, to debate on the topic of their effectiveness and, perhaps, get feedback or even some tips.

How we made Convead

When planning any new project, you have to find a compromise between your wishlist and available resources. In such moments, I always imagine a picture when on one shoulder sits the devil and whispers in my ear, "we Need more features! Nobody wants a service that can not do anything! Do steep front, people love with their eyes!", and another angel, and says: "We have limited money. Need to test the hypothesis with minimal cost. No one will die if the MVP will not export to Excel and it will podslushivaet layout!".

We decided to put ourselves a limit of 1.5 – 2 months to develop a prototype (aka MVP) and no more. In MVP decided to include only the minimum required:

the

gathering information about visitors and visits;
segmentation and Analytics.

Thus, in the current version Convead is able to collect and analyze information, but not how to influence the visitors. Export to Excel the way he did, the devil suggested :)

When we conceived Convead, we decided that focusing on the following business models for websites:

the

eCommerce
SaaS
Marketplace
Media

It is assumed that each connected site must specify the model for how it works. Why is it necessary to describe below. In the current beta version is only available in the model in eCommerce.

Each business model implies a certain life cycle of the website visitor. This lifecycle consists of a set of events that must happen, that the visitor has reached our desired result. In simplified form, the cycle can be represented as visiting certain pages. These pages we have called key pages, as they are the key to determine the key performance indicators of the website.

The business model for the connected site is determined not only a predefined set of key pages, but also which indicators we consider the key and what values of these parameters will be considered "good" or "bad".

For eCommerce model we consider the following key pages:

the

item Card (catalog, search, etc — any significant sign in page)
Basket
Checkout
order Confirmation

Scheme eCommerce site is presented below:

That is the scheme with the main parameters the user sees when each entry in Convead. Those who read books from the list at the end of the article, the scheme will seem familiar.

In the blue circles on top illustrates the distribution of the incoming traffic:

the

Return that we were able to return the result of a certain exposure.
Paid – paid sources.
Direct – direct calls.
Search – visits from search engines.
Viral – "viral" sources (yet simplistically believe that it is only social. network).

Together, these sources account for 100% of your visitors (Visits). The chart shows how the visitors move in our life cycle. At certain stages they fall off (leave the website) and it serves as a signal for the owner of the site that you want to change something in this part of the site.

The smaller the percent under the red label on the diagram the better. Bob will be easy to understand.

Next to the plan still have a simple sign with basic indicators:

the

% conversion
number of purchases
medium-check
total revenue

Key indicators can be tracked either for all traffic, or for specific segments (below), they are compared with the indicators of previous periods and determines the overall dynamics. The numbers vary not only in the table of the main indicators, but on the diagram of the life cycle. For example, you can see that in General you have the conversion rates are good, but among users of mobile devices very high failure rate on the cart page. This can be a signal that the page is something wrong in the mobile version of the website.
In General, we have great plans for the development of functional monitoring and lifecycle management of the site. For example, if you have a certain amount of representative data, we would not simply state the fact that you have a certain level of conversion, but also to compare it with the levels of conversion to other stores with similar parameters (topics, regionalism, etc.) and draw conclusions on how current performance is good or bad relative to the industry average.

Hence it is possible to affect the selected segment of visitors using one of the available channels of influence, by improving on it. Ideally we see it as: if something "bad", Convead paint this step in the diagram in red and will offer to change things for the better, by clicking on the button "Improve" and selecting the required effect (see effects below).

features in the current beta and future plans

Our idea was to give the opportunity to accumulate and analyze data (passive part), and then teach Convead to produce impact, to do something to change the situation (active part) and to extend the capabilities of analysts.

Therefore, MVP is able to a little while:

to Identify and recognize repeat visitors to the website, see their online activity and keep a history of interaction with the site.
to Count the income from each visitor.
to Segment visitors.
to Display metrics on key indicators of your site as all traffic, and segments.

A couple of screenshots of the interface (clickable):
the

the List (the selected segment of "With Revenue")

Card

future Plans

Effects

Convead was intended not only for analysts, and as a tool of active influence on site visitors, however, the main feature that impact in the current version not yet implemented. We will start to do it in the near future.

For a start, we decided to highlight the following impact channels:

Personalization: widgets, parts replacement, etc. page ("online influence").
Email message ("offline influence").

The idea is that with the help of offline influences we may try to return gone from the site visitors, and using online influences to motivate current active visitors to purchase.

All exposure pathways will support segments to be able to fully customize the effect.

Possible cases of application of the stress:

in a separate segment all visitors who were on site the last time a week ago, watched the product page, but didn't buy anything. Send them an e-mail "come Back, I forgive you!", the diagram of the life cycle to see the % of people who were able to return this way.
to Create an online impact that will modify h1 on the product page for that visitor, whom we know, writing there his name: "Arkady, well, buy iron Philips GC 8620, please!".
to Create an online influence, which will show the widget with a promo code for those visitors that fall into a certain segment (for example, they have a birthday today or they come to the site for the third time, but not buying anything).

an Automated collection of baseline data on visitors

Case studies for impacts you can think of a lot more, but, of course, the whole idea of targeted impacts of breaks on the impregnable wall of privacy of the visitor. After all, if we are not told anything about himself, we did nothing about it, by and large, do not know can not realize the most delicious cases of segmentation.
In this case, remains content segmentation by common values: an automatically determined location of the visitor, number of visits to the site, date of last visit, visiting certain key pages, etc.

The more data we have, the more flexible we can segment, the better to influence them. However, the collection of such data is a real headache for store owners. Our dream is to rid them from the need of gathering data about their visitors. Well, or at least to reduce the number of required actions to a minimum, passing only specific to your store settings.

We continue to think about how to legally collect the maximum amount data about the user, impeding that neither him or the store owner. Ideally, we would like to see receive all useful to segment data about the person (gender, age, marital status, work, interests, etc.) is automated as much as possible on the side Convead. Obviously, there are many obstacles and of course we know about them. But we have some ideas on this score and I hope we will be able to implement them. If anyone has good suggestions on this topic – Welkom in comments! :)

Integration

Oh, that magic word "Integration"! You do not have your API? Yes, you ~~the bottom~~ retarded service! Actually, by "integration" we mean not primarily its own public API (it will also), and the integration of Convead in most common CMS and SaaS services for eCommerce in the form of plugins.

We thought: why poor Bob needs to suffer and to configure life cycle of your store, the key of the page to ask someone to put the JavaScript snippet, if the vast majority of these VAS using some of the platform or the SaaS services? So, Bob, suffer you soon – we have included in our immediate plans to develop plug-ins for the most popular eCommerce solutions.

Improved features

We were in a hurry to release beta-version in set by us time, so some facilities still had to sacrifice.

In the near future, we plan to:

the

to enhance the capabilities of segmentation (to give the opportunity to create a chain of conditions);
to expand Analytics on Revenue (to accept not only the amount of the order, but the key parameters – the list of brands or categories of items in the order);
to add a beautiful (and for some stores, unfortunately, not very beautiful) trend graphs of key indicators.

Opinion

It so happened that in the countries of decaying capitalism today the eCommerce market developed much more than we do, and there are a lot of smart people who write interesting books on the subject of Internet marketing and so on. Us, for example, inspired mainly these books:

the

The Lean Startup: How Today's Entrepreneurs Use Continuous Innovation to Create Radically Successful Businesses by Eric Ries
Lean Analytics: Use Data to Build a Better Startup Faster (Lean Series) by Alistair Croll
Advanced Web Metrics with Google Analytics by Brian Clifton

Their way will be helpful to read for any owner of the online store. This, of course, is not a complete list of the great materials on this subject, I propose to Supplement it in the comments. It will be especially interesting to know if in your opinion something worthwhile on the subject of the Russian authors.

For us it was the first experience in the development of this kind of service, so want to say very much. Unfortunately, in one article is difficult to fit everything, so I decided to split the material into two parts.

In the next part (if the current part will be interesting) I'll do a story about the difficulties we face as took certain technical and organizational solutions. Will try to give as much delicious technical details of using Ruby-on-Rails, Postgres, Redis, Sidekiq, JavaScript, protezirovanie tables, load testing, etc. Until the idea to present it in the format "a description of the problem/situation is that the decision" to have the opportunity to listen to more experienced colleagues and to discuss this topic.
In fact, plans and ideas a lot more. Far more than our abilities. So I would be very grateful to those who try Convead and vote in our support-forum for this or that idea, or offer your own.

Today, the project enters the stage of closed beta testing. You may access it by leaving your email on the website http://convead.com in the appropriate form.

PS Actually, the article must have been published dlukyanov as the technical Manager of the project, but he do not have karma.

Article based on information from habrahabr.ru

Поиск по этому блогу

computer express