How institution botanists automate

Hi, Habr. It's been almost 2 years since my last post . I am very glad that he was useful for many. During these 2 years we have done a lot. And I would like to share my experience and impressions.

With 20 customer institutions from one country, we have grown to almost 4,000 from 12 countries. Our team has grown 5 times. Opened an office in Moscow and Warsaw. Released the application, in addition to iOs and Web, even on Android and Windows. Significantly increased functionality. Translated Poster into 5 languages. Over the last quarter, took part in 3 international exhibitions. We are close friends with such companies as Epson, Microsoft, Intel. We are closely watched by several large investors. The European company offered us a takeover.
')
At first glance it may seem that everything is simple and cloudless, but I assure you that it is not. The most interesting and dangerous only to be. In fact, we have passed only the first stage - the stage of testing the idea and building the first version of the product. Next is the most difficult stage - the stage of formation, when there is an internal optimization and restructuring of the company. That's what I want to tell you.

Small intro

Poster - cash register on the tablet. Allows you to automate the institution (cafe, bar, restaurant, shop ...) literally half an hour. Paid by subscription, which includes updates and technical support, so the cost of implementation is usually up to 10 times lower than that of similar stationary solutions (even without taking into account updates and technical support).

Technical support

This is our trump card. We work with HoReCa (hotels, restaurants, cafes), and this is a very conservative market. Especially in the CIS countries. If we take the European counterparts, then all the TP (technical support) from them is carried out by mail, at best, via chat. We immediately realized that we could not do without a phone and spent a lot of effort to build a hotline within the company. (I can tell in the comments about the pitfalls, if interested).

Initially, the TP we had just arranged. There was a responsible manager who connected new clients and then led already connected. Very quickly, we abandoned this structure, since upon reaching 20–30 paid clients per manager, its performance on new connections dropped sharply. This scheme is suitable for starting. But there is a risk that the manager will withdraw part of his clients upon dismissal. Clients get used to one manager, who often does not keep a history of communication with a client, this makes it difficult for a client to be assigned to a new manager.

The next logical step was the “conveyor”. We divided the department into Connection Managers, who helped clients familiarize themselves with the program and make the first payment and to Technical Specialists who solve current client problems. And it worked perfectly, as long as there were less than 10 such people, and no more than 1,000 customers. Further, it is difficult to scale such a structure. Customers are becoming more diverse and it is necessary to introduce an individual approach for each of them.

Now we have individual employees who specialize in the automation of coffee shops, during their work they helped open more than 100 such establishments. There are those who specialize in burgers, franchises or restaurant chains, food trucks, festivals ... This allows us to help our clients more accurately and quickly, because we already know all the subtleties of their business and help them figure it all out before questions arise. To build such a process took us a lot of effort and labor. Now we have an internal Wiki, where we have described all the processes of TP.

Tip : just need to get a Wiki, in which to describe the processes in the company and how to act in a given situation. This is necessary so that new employees do not step on the same rake, so that your decisions are clear and balanced, and do not depend on the mood and are a mystery to other employees.

Employee training

We pay special attention to training new employees. We did something like a mini-certification center. Each employee who comes to work goes through a week of training, at the end of which he makes a product presentation to the company. During training, he passes several tests on theory and practice. An experienced manager is assigned to him, who helps him to learn everything faster and is interested in successful training. This allowed us to put the process on stream and easily hire and train 2-4 new employees per month.

Tip : the process of training new employees should be done in advance. We have been postponing this moment for a long time and as a result have lost a lot of time. Read the Lean Startup! Would read before - would save a lot of time.

Technologies in TP

As the number of customers grows, the number of calls to the TP increases. Previously, we used regular mail, then began to not cope. Implemented Groove Ticket System. Integrated it with its CRM, now everything works like a clock.

We actively use chats in work. First tried Zopim . Cool, but expensive. Switched to SiteHeart (ideal for small teams and simple tasks). Half a year ago I discovered Intercom . Switched on him. First transferred registered users, then all. A couple of months ago, I noticed that competitors started using it too, which is a good sign.

For internal communication use Slack . In November, Mixpanel was introduced for detailed analytics. For tasks we use Trello . All this is tied together. Now, when there is an appeal from the client, the manager fixes it in our CRM, from where it falls into the tasks of the programmers and a message to the head of the department to monitor the status.

Tip : if the problem appears more than 2 times, you need to automate and describe the process.

Development

We have divided all development into project teams. Each project has a person in charge who leads the project from the beginning to the end and allocates resources. There is nothing unusual. Unless, there is a rule: once a month each developer should work one day together with technical support in order to better understand customers and their responsibility for the product.

Tip : everyone in the company should feel the users: their reviews, problems, needs. This is not a problem or manager’s responsibility; it’s every employee’s task.

Mentality

Recently, we often communicate with teams of similar products from Europe and America. We carefully study them at exhibitions. If you take the top 10 of these companies, then they are all ahead of us about sales. However, only a few of them have key functionality, such as inventory, costing, marketing campaigns, modifiers, fiscalization. It is simply impossible to work on the CIS market without it, because 90% of clients need these functions. In Europe, everything is easier. Customers are ready to compromise on the functionality, understanding that the cloud product is at the same time 10 times cheaper. In the CIS market is different. Clients in most cases are not ready to put up with the lack of the necessary key functionality, even taking into account the large difference in price, which makes the entry threshold for new teams large enough.

By the way, stationary systems often use this. They release a cloud version with bugs and poor functionality, and if a client comes to them and he is not able to pay $ 10,000 for their stationary product, they sell him their unstable cloud version. After a while, the client calls them, resents them for a long time, saying that "this is a cloud solution, and they are all so low cost, hence the price, and if he wants stability, then he will take a loan and buy a stationary solution." After such a client, it is already difficult to convince that only this company has such an unfinished product. The client is now sure that all cloud solutions work disgustingly.

Tip : test ideas. Go to the field and communicate with your customers. The experience of colleagues abroad can be deceptive.

API

Six months ago, we released an open API for developers. Now we transfer all our products to it. We have released an application for Poster Boss owners on it. Now we are integrating five independent projects. Several products have already been successfully integrated and soon we will tell about it. I believe that an open API is the key to success. Now, for example, two students from different countries have chosen Poster as their graduation project and are writing applications for it using our API. We even plan next year to make a thesis contest.

Tip : aim for openness. Share knowledge, practices. But do not release the API if you are not ready to support it and you do not have a person who will be responsible for it. API support and development is an ongoing process.

Seasonality

This fall we are faced with a very unexpected phenomenon - the seasonality of the business of our clients. Since Poster is distributed by subscription model, it is ideal for seasonal establishments that operate 4-6 months a year. We knew this, but we weren't ready for such a scale. As a result, the shutdown rate for October-November was 15 times higher than the norm. But it was literally two months and we quickly figured out this phenomenon. Now we consider these institutions a little differently in our growth model.

Tip : get ready for a spike. This could be a policy, a course or something else.

Expectations vs Reality

Here I would like to voice the most popular tips to articles about our product and how they did not pass the real test.

1. First of all, you need to release an application on Android, it is cheaper and more popular.

The most popular misconception. We launched Android almost a year ago and only 5% of our clients work on Android tablets. Only Chinese tablets with a disgusting touchscreen are constantly hanging and slower working cheaper than used iPad. If you want your business to work quickly, stably and without failures - you will have to take Samsung or Nexus. And here the difference is insignificant, but the WOW effect is present only with the iPad, for it is worth overpaying $ 50. We help businesses save $ 3,000-5,000, while saving $ 100 more on a stable tablet is impractical.

2. You do not have a warehouse / accounting / ... (substitute your own).

Now we have almost everything. When there was no warehouse - we thought that now we will make a warehouse and sales will go. It was the same with accounting. But the reality is that more than half of customers use only 20% of the functions. In fact, less than half of customers use warehouse and technical cards from the beginning of using the system. When we did this analysis, it was a shock for us. We are currently working on gradual customer training in basic functions in order to increase the number of those who use most of the opportunities.

3. Tablet each waiter / client in hand.

Less than 20% of our customers work with more than one terminal at the establishment. Of course, this is economically advantageous: the waiter does not rush between the hall and the kitchen, the load on it decreases, in the end, you need fewer waiters. But in reality, only a few owners calculate it and implement it.

4. A beautiful terminal interface is most important.

The system is bought by the owner, not the waiter. In most cases, the owner doesn’t care what the terminal looks like for the waiter. Yes, and the waiter does not care. The main thing is that the terminal worked quickly, the waiter could serve more guests and earn more tips. Only 25% of customers add photos to all products on the tablet. And about the same does not add a single photo of their products.

5. Subscription puts the client on the hook, forcing you to constantly pay.

Another popular misconception. On the "hook" it sits down stationary systems. After all, after he paid $ 5000-7000 for the equipment and system, he has almost no chance to refuse or return them. You have to pay for updates and technical support and wait for her for hours. With us, a client can at any time export all data, just stop paying and switch to another system. He knows this, we know it. Therefore, we are doing everything possible to ensure that the client is always satisfied and that he does not have such delusions.

Tip : Learn how customers use your product and help them better integrate with it. Do not believe what is written in the comments. Believe only numbers.

During this year we faced a huge amount of problems. Many we managed to solve, many still to come. The most important thing that I understood is that business is a process. This is not something finite. You can't just solve a problem. After a while, it will change and the solution will no longer work. We have to invent something new and optimize. You can not stop and just watch.

It would be interesting to hear who came across similar things and how did you solve them?

PS I will be glad to answer questions in the comments.
PPS We are looking for interesting projects in the restaurant industry to integrate with our API. Who cares - write in the LAN.