How we started to launch our SaaS product to the global market

The attentive reader of our blog remembers that the whole company began with the idea of ​​creating an automation system for the service business (IT outsourcing, b2b Internet services, and later realized that internal IT services have the same needs), which will be distributed through the SaaS model.

The same SaaS, which we called SmartNUT, is now in a phase of closed operation. This means that there is no beautiful site yet, but anyone who wants to connect can submit an application through a google form on the stub page (on the same stub page, you can briefly read about what will be in SmartNUT). So, since there is no site and the service is not running in the official production - we had no choice but to go to the global market. This will be further discussed.

As a spoiler: the discussion further deals with the use of contextual advertising Google AdWords on the global Internet. And as a disclaimer: we understand that this is not the only way to find potential customers in the global market (and, possibly, not the most effective one - we cannot yet judge other ways). Well, and more - we are never a guru in terms of contextual advertising management.

First shot: quickly look at the possibilities

One fine day in July it was thought, why not look at the effectiveness of Google contextual advertising in terms of attracting potential customers for SmartNUT in the global market. And so itched my hands that I wanted to see some result quickly and cheaply. In the end, did this:

• The English-language blank page , which briefly explained what SmartNUT is and honestly reported that the system is not yet available, but you can leave your email to inform you about the launch (15 minutes);
• A form in Google, which could be accessed from the English page-stub in order to leave your email and a little more information (5 minutes);
• In AdWords, one ad was created using helpdesk, servicedesk, etc. (one of the main functions of SmartNUT);
• In AdWords threw 4,000 rubles;
• Put a daily budget limit of 1500 rubles.

As a result, we got the following:

• Google suddenly wrote off 18% of VAT, so in fact we have ~ 3400 rubles left;
• For 2 days with a little, 311 clicks were received with an average cost of ~ 11 rubles (if you don’t forget about 18% deducted earlier, then the average cost of a click is ~ 13 rubles);
• Out of 311 clicks, 11 registrations were received (conversion ~ 3.5%, registration cost including VAT deducted ~ 363 rubles).

It should be noted that it is pointless to conclude that it is expensive or cheap at the current stage - for the conversion of registered users into paid users is still unknown. But if you look at the average 3-4% for domestic SaaSs, then the cost of the attracted client would be ~ 9-12 thousand rubles, which is a lot. But rather - on the verge of payback, taking into account the average lifetime of the client and the average cost of systems of this class. However, I don’t want to go into the metrics in the framework of this post (and those who wish can read the recording of a wonderful story on the subject)
')
In addition, no optimization of contextual advertising has not yet been carried out, and visitors are “lend themselves” to a page that was not made in the best possible way. In addition, we received a certain number of registrations, especially targeted (judging by the sites of companies whose representatives left applications). So you can dig further.

Second shot: normal page

Then we decided to see how the conversion from visitors to registration will change if there is a normal landing page. Freelancer made a new one . And again they threw 4000 in AdWords to show the same ad (the 18% VAT write off was no surprise to us). The results are as follows:

• 251 clicks in two days - it’s not difficult to calculate that the cost of a click has increased to ~ 16 rubles;
• 251 clicks resulted in the same number of registrations: 11 pieces.

In other words, despite the fact that the conversion increased from 3.5% to 4.3% (although the difference can be a measurement error on such a number of transitions), the cost of a click also increased. As a result, the cost of registration remained the same - 363 rubles.

As a result of the manipulations, we spent 8,000 rubles, receiving 22 registrations with an average cost of 363 rubles. We decided to look at the geography of registrations:

• India: 3 registrations;
• Mexico: 2 registrations;
• China, USA, Pakistan, Poland, Romania, Egypt, Indonesia, Morocco, Peru, Nigeria, Ghana, Cote d'Ivoire, Spain, Jordan, Dominican Republic, Ecuador: one registration.

And one registration with this data: هحجنتمربغ عامح

Google translator suggested that it was Urdu (but could not translate), whose carriers live in India and Pakistan. Therefore, we did not take this registration into consideration.

The next sensible step: to combine the geography of registrations with the geography of clicks in order to identify the most interested countries and target the ad for them.

It would be reasonable if the number of visits from any country was sufficient to build statistics. More or less the number of visits was from India: 30 pieces. But this is not enough to draw conclusions about the real conversion.

Therefore, we decided to go through the BRIC countries (excluding P from there - in Russia and the CIS we will start next month after the public release; and the flow of leads from the community of IT company leaders we have created is enough to run the system).

Third shot: geo-targeting

"Divide" in AdWords old ad into three countries (Brazil, India, China). Fill in each ad for 2000 rubles. We look at the disappearing 18% VAT and wait for the day (in reality, in each country, the money disappeared in 2-3 hours). Daily budget constraint removed.

Total:

• China: 0 registrations (sadness, sadness);
• India: 5 registrations, 150 clicks;
• Brazil: 7 registrations, 93 clicks.

Thus, India showed close to the world average conversion and registration cost (400 rubles for registration), Brazil took the lead (285 rubles for registration).

At this experiment, they decided to stop before the appearance of the English version of the service (as it is known that registrations, if they do not carry out any work, become “dead” over time).

Morality

For ourselves, we concluded that contextual advertising on the global Google-e can serve as a normal source of leads (we don’t say about the quality of the leads - until you make a sale, it will not become clear). The cost of registrations turned out not exorbitant. There is a feeling that it can be reduced by working on the optimization of the advertising campaign (including with the involvement of people with experience).

ps Those who want to learn more about SmartNUT without connecting to the operation can read and see screenshots here .