How we conducted interviews with the first users and what came of it

    There is nothing more important in business than understanding the client and his needs. Sam Walton in his book “Made in America. How I Created Wal-Mart ”said that a deep understanding of the client and his needs is the key to a successful business. We have seen this from our own experience when communicating with customers at Teamdesk .


    About 7 years ago, when I created the first Internet projects, I didn’t even think about my audience - for me people were faceless traffic and just numbers. A thousand more, a thousand less - it did not matter. I didn’t care who these people were, then it seemed not important. I realized the whole depth of my errors only when working on my first b2b company: for several months we struggled unsuccessfully to find channels to attract customers. But how to attract them if you don’t know anything about them ?!

    Then someone on top took pity on me and sent Steve Blank's book on Customer Development. Only then did I realize that I knew absolutely nothing about our customers. At that moment, in order to move forward, we had to take a step back. Then, to correct this misunderstanding, I traveled to clients and conducted interviews, talked with managers, sat behind employees and watched how they work in the system. That was what I needed.

    Wise by the gray hairs of our previous experience, at Teamdesk we immediately set ourselves the task of learning as early as possible and as much as possible about our customers. Despite the fact that I understand the subject area quite well, neither I nor my partner knew anything about our target audience. Who uses ticket systems? How many people are usually in support? What functionality is needed first? We had only hypotheses. We only had to answer all these questions.

    What are we interested to know about customers?

    We needed to somehow structure our knowledge about customers, so we sat down and together with the guys identified the main groups of information. We wanted to know all (or almost all) of the users of our Help Desk system. Of course, the maiden name of the head’s mother will be superfluous, but at least we should clearly understand the following important points:

    • The general profile is what kind of company it is and what it does, where it is geographically located and how many employees it has in general and in particular the industry is in support.
    • Problems / pain - what kind of pain the client has, what disturbs him and does not suit him in the current process of customer support, why he came to us, what results he expects from the system.
    • The decision maker, the budget holder and the end user - we try to understand who in the company decides to purchase, who conditionally signs an account, who will use the system later. Somewhere it is one person, and somewhere 3 different.
    • The decision-making process and the purchase process - we also try to understand how the system chooses inside the client, who discusses the purchase and how, how the testing happens.
    • One day in the client’s life - we want to know how our client spends his working day, what programs he uses, how much time he devotes to the responses in the ticket system, which devices he uses for this.

    Here in the aforementioned Steve Blanca book on Customer interviews, a total of more than 40 questions for customers are listed. We all made it a little easier.

    What did we ask?

    When it became clear what information is needed, we formulated a total of 12 questions. Where it was possible, we tried not to ask unnecessary questions, but to seek information in open sources. For example, if we know a domain and see a typical online store on it, then why ask something else, if it is already clear that this is eCommerce. Here is the list:


    In addition to the list of questions, we also decided to look at the sites of competitors. Usually there are sections “our customers”, “for whom” or “success stories”. All this allows us to understand who is already using ticket systems (albeit not ours yet). As a result, even before the survey, we had some idea who we would be talking to.

    Also for ourselves, we formulated these simple rules for conducting a survey:

    • We always ask everyone the same questions. Once I tried to communicate with clients in a free conversation format and to outline all that was said. When more than 10 such interviews accumulated, it was extremely difficult to draw practical conclusions - all the information was fragmented and poorly structured.
    • In no case do we sell - there was such a sin behind us, when we gradually moved from questions to sales. Clients closed and dialogue did not work.
    • We speak 20%, and we listen to 80%. Usually, after a brief introduction, we ask the first question and carefully listen to and outline the answer. Let us talk to the maximum.
    • If they spoke less than 10 minutes, then the interview failed. People usually willingly share their problems and talk about their work. 10 minutes is a minimum.
    • We prepare questionnaires in advance with questions. It is very convenient to simply take a sheet and write down a specific answer to the question there. Then put these sheets in a pile and at once look at 20 answers to 1 question.

    Well, there are questions, the first information was received, it was time to take up the phone / skype and start communication ...

    How did we ask?

    The first pancakes were lumpy - the dialogues did not stick together, customers did not want to waste time and communicate. We are a little upset. Then they realized their mistake: it was necessary to prepare and rehearse the conversation script in advance and, most importantly, the conversation should not be intrusive and strain the client. Otherwise, the dialogue will not work, and the answers will be in the format "Yes", "No", "I don’t know."

    Then we sat down and sketched the following script:

    “Good afternoon, this is Nikolai from Teamdesk. Do you have a minute? ” - If the client gave an affirmative answer, then we continued.

    “I would like to consult with you on working with the ticket system. We conduct a small customer survey to better understand your tasks and needs. For us it is very important. Takes no more than 5 minutes. Do you mind? ” - Then followed the questions on the list. Sometimes we did not keep order.

    Thus, we greatly increased the responsiveness of customers and decided 3 important points:
    1. clarified whether it is convenient for the client to talk. If not, then they called back at the agreed time.
    2. clearly defined the purpose of communication, people began to understand what they were wasting their time on.
    3. designated a time frame of 5 minutes and showed the importance of this pig for us.
    4. A personal meeting is certainly better, especially if you work with corporate clients. But we usually used a phone or skype. The survey was carried out either immediately after registration in the system, or when they saw that the user had configured the system and actively used it.

    This is how one of our real profiles looked like with short notes-answers (I won’t, of course, name the company, all names have been changed).


    And the main question: when to stop the interview? Never! We will continue interviews and communication with clients. This process should be continuous, as the market may change, the product will change, new customer segments may appear.

    How did we analyze the questionnaires and what conclusions did we draw?

    When we had accumulated the first 10-15 questionnaires, we could already draw the first conclusions. Some of them confirmed our hypotheses by segment and price; some of the answers gave new insights.

    We realized that some of the clients are looking for not a help desk for customer support, but a service desk for automating the internal IT service. These are not our clients, since they need a completely different functionality and they have everything in a different way: decision-makers, budgets, etc. It is possible that our system is not accurately described for whom and for what our system is, maybe people just don’t know the difference between a help desk for customer support and an IT service desk. We'll figure out!

    We also realized that many users are comfortable paying 1000-1500 rubles / month. for 2-3 employees, this means that our price hypotheses and current tariffs have hit the mark. Sales will show how much. The hypothesis with the "Webmasters" segment did not materialize. There was an assumption that they needed a ticket system to support site visitors, but this is not so. They have no interest (and therefore problems) in this.

    Communicating with customers, we began to better understand what functionality they need right now, and what can wait. As a result, we slightly adjusted our roadmap. Also, lively communication with some customers allowed us to build long-term and friendly relationships, and it costs a lot.

    Friends, I hope our experience was useful to you and will allow you to better know your customers. Subscribe to our blog on Megamind and ask questions in the comments.

    Also popular now: