From 0 to 1: Finding a Pilot Customer for an AI NPM Startup




Early-stage marketing



In this case study, we share how we helped a client find the first enterprise customer to do beta testing of a new machine learning (ML)-based network performance monitoring (NPM) tool.

agreement on cooperation

Why only one client and not 5, 10, or 100? Implementation and tuning of the tool over such a large network would require far greater resources that were available to our small bootstrapped team, who were doing it for the first time. We wanted to get it right; otherwise, what’s the point?

As part of the process, the client was also able to develop a clearer understanding of who their target audience was; we launched a website for them and developed a content plan.

Read on to find out:

  • how we defined the ideal customer profile (ICP) and buyer personas;
  • how we ‘hacked’ Quora to help establish the strengths and weaknesses of competitors’ products;
  • how we secured that first beta-testing customer.

What is NetSequency?

NetSequency is a young startup that provides NPM software powered by ML. The software assesses a massive amount of data autonomously, in real time, with zero human effort; it also shows correlated network anomalies and can even find the root causes of them.

The target audience for this tool are companies that manage complex networks with hundreds and thousands of network devices: enterprises and managed service providers (MSPs).

For enterprises and MSPs, this software can augment the existing monitoring tool or be used as the primary NPM, significantly decreasing workload for network engineers.

main page of NetSequency's website


NetSequency is a startup that has a very valuable solution to a complex problem. The objective was to find the first customer to beta test their product. They needed someone who has a large complex network and is ready to deploy their network collecting engine to send network data to the ML engine. It turned out to be problematic to find a customer who would test the product in their day-to-day work environment using real data, not least due to concerns about data confidentiality.

Consequently, NetSequency did not exactly know who to approach with the offer to test, or how to talk about the problem that the product could solve.


  1. The market is saturated with solutions. The main players in the niche have established a reputation and gained trust from numerous customers. SolarWinds MSP has been on the market since 1999 (source), Zabbix since 2005 (source), and Auvik since 2011 (source), to name but three.
  2. There’s also fierce competition between AI tools in the market, tools such as Kentik and Anodot – with total funding of $61.7 million (source) and $64.6 million (source) respectively.
  3. The target audience has already found an adequate solution and they aren’t aware that there is a superior way to fix the problem; this means they aren’t spending time making direct search requests on Google.

The Services We Provided

An Essential Website

Although the initial goal was to sign up a customer to beta test the product, several key steps had to be taken before launching the campaign to find a customer. 

First, we had to launch a website for the product and identify the target audience. Although a website may seem to have been separate to our central task, we quickly discovered that it was crucial:

  1. Almost everyone we talked to asked about a website.
  2. We were losing the trust of the potential customers. A change in tone was apparent when they discovered we didn’t have a website to direct them to.
  3. Some activities we planned required a landing page with a registration form.

It made sense to launch a simple temporary website. So that’s what we did!


Research of the target audience was one of the customer’s central requirements, as: 

  • the market is relatively saturated;
  • they were looking for an independent point of view on their positioning;
  • we needed to understand how potential NetSequency clients would go about choosing a product, what their goals were, and what problems they regularly faced.


These were the primary steps in our project:

  • Target audience research:
    • create ICP and personas within different ICPs;
    • research the demand: find out how customers solve the problem and document the buyer’s journey;
    • identify the sources of information and communication for the target audience;
    • interview target audience: we went to the relevant communities on social media communities and talked to potential customers.

  • Website:
    • deliver the initial version of the website;
    • analytics and lead capture setup;
    • content plan and a blog article.

  • Outreach campaign:
    • lead generation using different channels.

We split the project into four stages:

  1. Research.
  2. Setup.
  3. Launch.
  4. Results.

The Research Stage

Product Review and Target Audience Research

We started with product features and the value they could offer customers. It was an essential step in defining the target audience and informing the future copy of the website.

Another essential part was to identify competitors and their position in the market. For example, we needed to determine what they were offering, who they were targeting, and the platforms they were using. 

Looking up features of the competitors’ products helped us get a better understanding of where NetSequency could stand among the competitors.

We drafted a value proposition for NetSequency and identified who we could verify the value with, and where it could be done.

valuable features of NetSequency

We grouped the features into those that are provided by all tools on the market and those specific to ML/AI products. Then we narrowed down the list to three valuable features of NetSequency and compared them to those of competitors.

How to Learn What our Competitors Think about Other Products’ Strong and Weak Points

To better understand the features of the products on the market and the extent to which they vary, we turned to Quora.

It can sometimes be difficult to get noticed on Quora. Luckily for us, after posting a few questions talking about our competitors’ products, we started receiving valuable comments from people inside the industry. In fact, an ex-CTO of Kentik, whose product we had mentioned, explained what they felt Kentik’s advantages were.

relevant post on Quora

Compiling a List of Relevant Communities

To find out what our target audience was thinking, we had to find active communities where network experts discuss industry-related issues. Participating in the right communities is the best way to: 

  • discover current, discussions and trends;
  • distribute content;
  • source data for scraping for further activities;
  • advertise;
  • make announcements, such as previewing an upcoming article.
spreadsheet with relevant platforms for NetSequency

Finding Actual People on LinkedIn that Match the Defined Personas

We listed individuals who may be interested in using NetSequency and sorted them by place of work and position. Here are a few reasons why:

  1. First, we needed to verify that the personas we created were correct. Instead of showing hypothetical portraits, we offered the client to look at the profiles of real people.
  2. This list could be used for future outreach campaigns.
  3. It helped us to build the audience for ad targeting.
spreadsheet with matched personas on LinkedIn

How to Interview Your Target Audience to Get the Answers You Need

Our next step was to talk to network engineers to help us better understand the challenges they face and how NetSequency could help.

We prepared a list of questions for the interview with people who matched the decision-maker profile in the relevant field. Below we talk more about how we came up with these questions:

a list of questions for the interview

Questions to Ask in the Discovery Interview

Here’s how we shaped our questions to get the information we needed:

  • Learning about each individual’s background
    • We wanted to make sure that the interviewee’s experiences were relevant to us, especially as we were paying for their time.
  • Learning about their sources of information
    • We asked where they got their information from, regarding new developments in their field, new tools, and other updates from peers. We could then target these platforms as both places to gain further insights and also post ads or other content.
  • Decision process
    • We wanted to learn more about the general decision-making processes in companies the interviewee has had experience with.
  • Switching to a different solution
    • We needed to understand the obstacles to decision-makers switching from one solution to another. 
  •  Goals and obstacles of the working process
    • With this series of questions, we wanted to learn what they wanted to achieve. This would help us pitch NetSequency more precisely to potential customers, for example by emphasizing the potential improvements in MTTR time. 
    • We wanted to find out as much as we could about recurring problems that each interviewee faced.
  • AI
    • Since the AI component is one of the main features of our product, we wanted to find what experts in the field thought about it.

We identified Reddit and other social media platforms that were most frequently used by the target audience. 

We sought responses from people in the sector who were not employees of the direct competitors of NetSequency. 

We offered a $50 Amazon card in exchange for twenty minutes of their time, making it clear that we weren’t trying to sell them anything. This incentive was an attempt to show that we had serious intentions; although, offering a monetary incentive can be a double-edged sword: it’s great motivation but there is a risk that it will attract the wrong audience. 

relevant post on Reddit

Posts like this often don’t get enough relevant answers on Reddit. We were fortunate enough to receive plenty of valuable comments and private messages from members of our target audience.

We also reached out to network engineers from our own professional network and the insights we obtained agreed with the feedback from the social media platforms.

Three significant findings of our research were:

  1. We discovered that the decision process in enterprise contexts was significantly longer and more complicated than we had expected.
  2. NetSequency would likely benefit from being positioned as an auxiliary tool rather than a replacement.
  3. The majority of respondents treated the AI component of the product with skepticism.

The Ideal Customer Profile

An ICP helps define a target for a company’s messaging. It helps determine the content of the message, the target audience, and the appropriate platforms to use.

Startups without marketing people on their team tend to skip important steps such as market research, identifying a target audience, and experimenting with product positioning. 

This leads to diluted messaging from the brand, which is then directed to the wrong audience, all leading to poor sales. This is why we always ask our customers if they know who their target audience is. If they don’t, we carry out research, allowing them to save time and money while ending up with the satisfying feeling of pitching to people who are actually interested in your product.

In the case of NetSequency, we used the results of the research to create an ICP for each persona. The insights from the customer interviews helped us verify the value proposition and customer profiles and understand where NetSequency could sit within the market.

We knew that NetSequency’s solution was aimed at enterprises. Therefore, we narrowed down what enterprises fit the ideal customer profile. Then, we created a second profile, the MSP.

ICP characteristics

Since you can’t pitch to an enterprise, but to its employees, we further identified the personas’ personal profiles.

ICPs main profiles

Multiple Useful Lists

By the end of our research stage, we had established the ICP. One of the most important subsequent points was to find out where the target audience consumed content and met peers online.

Knowing the platforms helps us find out about trends in the industry and allows us to observe both recurring and novel challenges faced by the target audience. Reading through the conversations on the various platforms, we were able to get a better understanding of how to “speak the same language” as our target audience. 

Furthermore, we analyzed these websites regarding their potential as content distribution platforms: we checked if they accept guest posting, the size of their audience, topics commonly covered, amount of comments, likes and votes, and other relevant activity. 

Below is a list of relevant platforms for NetSequency.

spreadsheet with relevant platforms for NetSequency

Analyzing Competitors’ Content

By analyzing competitors and their content, we learned that they all invest a significant amount of time, money, and effort into their content strategy. They produce a lot of quality content.

results of NetSequency competitor's analysis

The Setup Stage

Wireframes and Illustration

Now that we had learned about the target audience and the product, it was time to launch a website for NetSequency.

We used Figma to draw wireframes, approved them with NetSequency, then we hired a designer to prepare the illustrations and website mockups.

Website setup

We set up WordPress with all the necessary plugins. 

We used the information we gathered from the research stage to create copy aimed squarely at the relevant audience. We identified that the AI/ML component could be the line that would help NetSequency stand out among other tools on the market.

With NetSequency, we worked on the website in iterations, presenting drafts and making revisions according to feedback. We repeated this process until we were satisfied with the final look.

main page of NetSequency's website

We prepared some attractive images to illustrate key features of NetSequency.

illustration of key features
illustrated key features of NetSequency
key features of NetSequency
illustration of NetSequency features

Website Analytics and Lead Generation

Our next step was to ensure that we tracked visitors to the website, so we connected Google Tag Manager, Google Analytics, Google Search Console, and HotJar. We set up events, tracking, and goals for the website.

Then we created an email account. We did research on ActiveCampaign and MailChimp. We used these platforms to collect email addresses through a form, for further email marketing activities.

The Launch Stage

Now that we knew what audience to target and we had a functioning website, it was time to get the first beta customer on board. We prepared an outreach campaign and gathered contacts who matched the ICP.

To maximize takeup, we heavily personalized the messages we were sending; we didn’t want them to look like spam. We looked through each individual’s profile and got to know them by reading their recent posts or comments, so as to better craft a genuinely personalized message. We did not send the same message across the whole list of contacts. It took time, but the results suggest that it paid off.  

For even greater personalization, we sent follow-up messages too; this put prospects at ease as we further demonstrated that our messages weren’t spam.

We ended up with three potential representatives who were ready to beta test NetSequency. Our customer paused the two that were less relevant and picked the most relevant one.

The Results Stage

Despite occupying a challenging niche with tough competition, we managed to find the perfect fit to beta test NetSequency’s product. The customer was really excited to begin the testing process and was ready to become the first paying customer if the solution fits their requirements.

While most of our clients are found through referral, NetSequency found us through recommendations on Upwork. Here’s the rating NetSequency’s founder left on Upwork:

job feedback on Upwork


  • Found the first enterprise customer to do beta testing of a new ML-based NPM tool.
  • ‘Hacked’ Quora to help establish the strengths and weaknesses of competitors’ products.



Need a hand finding your first customers?