How to Launch Research Reports

Go back


How to Launch Research Reports

(Or how we released 18 Labs reports in 18 weeks)

  • This playbook focuses on building and operationalizing content production and how HockeyStack streamlined its Labs production.
  • With this playbook, you can learn how to build a content engine using your own first-party data and publish authentic thought leadership content for your audience at scale.

Summary 

We’ve published 18 HockeyStack Labs reports in 18 weeks in 2024, generating nearly 100K unique visitors and millions in pipeline. HockeyStack Labs has become our golden goose—more than half of our pipeline comes from these reports. Our website traffic has more than tripled, and these reports have greatly accelerated our sales cycles.

However, it wasn’t easy, and we had to learn a lot to scale this process. As we kept publishing reports, more and more people started to message us about how we do it, and these messages inspired this very first playbook.

In this playbook, we’ll cover everything we’ve been through and how we managed to scale the Labs process.

The most authentic thing you have is your data. Other companies can copy your structure or processes, but they can never replicate your data and how you present it.

Using your own data as thought leadership content puts you in a unique position among your ICP. These reports are not just for attracting new customers; they are also valuable for your open deals and existing customers. By providing truly unique and educational content, you build unmatched trust.

Additionally:

  • It brings high-quality traffic to your website.
  • It gives you free brand visibility as other people share these reports when you present unique data points.


Step-by-Step Guide: How to create your first report

Some might argue that you need a scalable concept and should plan everything in advance, but my recommendation is the opposite. I suggest treating this project as a one-off initially because you have no idea if it will work for you. If you spend too much time planning and investing resources, and it doesn’t work, it might cause more harm than good.

HockeyStack Labs reports worked really well for us, but this might not be a recipe for success for everyone. I believe they worked for HockeyStack for a couple of reasons:

Organic advantage: We had data from many B2B SaaS companies that we could easily anonymize and analyze. No major company had this—not Linkedin, not Salesforce, nobody. Smaller companies had no idea how to use or analyze their data.

Limited competition: There were not enough reports in the market that our ICP could read. The ones from Gartner or Forrester were mostly geared towards enterprises, not companies with 100-5k employees. However, you need to think strategically here: why was there no report? Is it because nobody can generate one, or is it because there is no demand? This is such a crucial difference.

No recent data: Even among the available reports, there were only a handful of recent ones with limited datasets. As metrics change significantly every quarter in B2B SaaS, reports from two or three years ago were mostly irrelevant.

Understanding the ICP: I was my own ICP, so I knew exactly what data points my ICP needed to see and what they would be curious about.

*If you are uncertain about your ICP, you might need a subject matter expert. There's nothing wrong with this; you may understand your ICP and know their pain points to some extent, but may not fully grasp it. For instance, if you target the procurement persona, you can only empathize with them up to a certain point, unless you’ve been marketing to them for years. Therefore, collaborating with a subject matter expert in creating these reports could be beneficial.

At the end, you should be able to say: “My ICP would need this data as it would make their lives easier.”

I believe the points mentioned above are the main reasons why the Labs worked so well, but until we launched the first report, we had no idea if it would be successful. These reports were one of our experiments, so the initial investment was minimal.

Success metrics

Since this was just one of our many experiments, we were conservative about what success would look like for us. We didn’t want to bet on pipeline or closed-won deals; rather, we focused on vanity metrics like engagement, comments, and website visits.

After seeing that our ICP was genuinely interested in the first report, we decided to continue with the next one. The Recap 2023 report became the most visited pages on the website that month, the bounce rate was low, and the time on page was more than 3 minutes. This clearly indicated that our ICP found the Labs content helpful.

However, we didn’t have a clear process up until the fourth report. This had pros and cons.

Pros: We avoided wasting time on long-term planning, success metrics, or budgeting.

Cons: We spent significant time in structuring data processes, and decided content topics weekly.

Even now, we don’t have a strict process for choosing topics. I choose the topics myself and only write about what I genuinely find interesting; I’m fortunate to be my own ICP. But if i wasn't my own ICP, I'd brainstorm with a subject matter expert or a product power user to understand what would really interest them.

For example, if you are selling to developers and you’re a marketer, the easiest way is to talk with your customers who use your product daily.

Once you have some answers, it’s easier to reverse engineer them to find ideas. For instance, if you’re selling an AI B2C chatbot for customer service directors, and your ICP is customer success directors and above at giant corporations like Lufthansa, Delta, or United, you need to tailor your content to their specific needs and challenges.

If you already have some power users, I recommend reaching out to them directly with utmost personalization. You might consider offering them a discount on your product, unlocking a new feature, or giving more credits.

For example, if they mention, "I don't know how long it takes other companies to solve their customer queries in AI chatbots," or "I don't know how many chats are handled by AI," you can work on a report comparing customer query times across your customers. This data is likely available in your product but might not be easy to find or export, so you’ll likely need a developer or someone with intermediate SQL knowledge.

In our case, Kris took some SQL courses on Udemy, which was enough for him to understand the fundamentals. Then our CTO, Arda, trained him for a few hours. Our plan was to hire an SQL developer for these queries if Labs worked out.

For our first report, we decided to start with the 2023 benchmark data since 2023 had just ended, and as a marketer, I knew other marketers would be interested in that information. I listed out the data points we needed, then Arda and Kris had a call to discuss which data points were possible to retrieve and if we could create a full report with those data points.

Since we didn’t have a process in place, it took Kris a week to gather all this data, followed by a few days for me to organize the data and write the Python scripts to analyze it.

This is our current Labs process:

1. Find the idea for your report
2. List down the data points you will need.
3. Discuss with the team the possibility of getting these data points.
  • For example, you might not get the exact data of MQL:CW close time, but if you can get the data of MQL creation date and deal closed date, you can still calculate this.
4. Identify the closest data points if you can’t get the exact ones.
  • For example, we couldn’t find the data for outbound deals with marketing touches in the last 14 days before the first sales outreach. Hence, we decided to focus only on LinkedIn touches. If you can still create a report with the available data, that should be sufficient.
5. Get the data, categorize it.
  • Once you have the data, the next step is to categorize it and work on a structure. How are you going to present it? How will you structure your report?
  • Since I already knew how to use Python to analyze data, we didn’t need external help. If you have someone on your team with Python or data analysis skills, then perfect. If not, start with easy and simple data points that you can analyze yourself on spreadsheets with pivot tables.
6. Ready, set, go.
  • Once you analyze the data and structure how to present it, the next step is to write about that data. I don’t have any specific tips here; I just look at the data and start writing.
7. Distribution 
  • You might have written great content, but if you can’t distribute it effectively, it might all be for nothing. We start our promotion with Linkedin, where the company page and everyone at the company post from their profiles.

  • We realized that reach decreases if everyone at the same company shares the same content at the same time. So, 25% of the company posts it on Tuesday, and the rest post on Wednesday, Thursday, and Friday, respectively.

  • We created an email newsletter by using an automated Intercom chatbot that shows the Labs Newsletter sign-up form if someone spends more than 30 seconds on Labs content. This helped us build a newsletter audience of 15k people. Now, as soon as we publish new Labs content, we also send out the newsletter.

  • For the first three reports, we sponsored marketing influencers for distribution. This was a beneficial move for us as it gave the Labs reports a lot of visibility. Once we gained that visibility, we didn’t need influencer boosts anymore. In fact, after a while, these influencers started posting about our reports for free because they found them interesting. That was the biggest indicator of success for us.

8. Repurposing 
  • You can’t just publish and forget; you need to use it repeatedly if the data you provide is unique.
  • Currently, we repurpose our content as follows:
    • P = Report published and shared
      • (Company page, employee profiles, email newsletter)
    • P+2 = Another company page post; carousel with important data points.
    • P+4 = Video Snippet
      • (Youtube Shorts, and Linkedin)
    • P+7 = Linkedin Newsletter
    • P+10 = Video
      • (Youtube and Linkedin)
    • P+13 = Second Snippet
      • (Youtube and Linkedin)

Operationalizing the Process

To scale, the first question you need to ask is, 'What's taking the most time in this process?'

In our case, it was gathering the data. We spent days getting some data points, but moving forward we had to be faster. So we started asking some questions; is there a way to get that data faster? how can we reuse this data in another report?

  • One of the things we realized in our case was that if we unified the property names across our customers, it would significantly reduce Kris’ workload. At that point, Kris had to find the definitions manually - for instance, each company had a different definition for a deal, some might call it SQL, another SQO, another Opp, and so on. To address this, we built something internally to unify all these definitions; it worked so well that it also solved our need for an SQL developer.

With the fourth report, we realized that Labs was going well, and we needed a game plan to operationalize this process.

Things to consider

  • Once you agree on the time period between reports, plan for an extra topic just in case you can’t get a data point for one of the reports.
  • Always include a methodology and explain how you obtained and defined the data; otherwise, it might lead to disputes. I learned this the hard way. There will always be someone who accuses you of presenting the wrong data.
  • Always verify your ideas with another peer. Just because you find a topic interesting doesn’t mean everyone will.

Before the launch


  • Does the way you presented your data make sense?
  • Did you leave out any data points?
  • Does your report have a storyline, or did you just dump your data?
  • Have you checked the website URL? Is it working?
  • Are there any typos in the report?

Summary

  • In this playbook, we’ve covered our process of creating and publishing the Labs reports. It might seem challenging at first, and you might end up asking for a lot of favors from your technical team, but if you make it work, it’ll all be worth it.
  • Once you publish reports, you can easily repurpose them as videos, podcasts, social media posts, and sales collateral. In a way, you can create your own ReportVerse just like we’re doing with LabsVerse.

Written By
Canberk Beker
Head of Growth at HockeyStack