Authors: John & Yolo
Translated by: Aaron Huang
On June 20th, Slack went public with a pop. Trading on the New York Stock Exchange, the stock opened at $38.5 per share – 48% above its reference price and giving it a market cap of $23B. It’s hard to imagine that it was only 5 years ago when Slack launched its first product. Confident in its cashflows, Slack also surprised the public by following Spotify’s path of going public through a direct listing (DPO) rather than the traditional IPO route. Today, Slack boasts 10 million DAU and 600,000 organizations as customers, including 77 Fortune 100 companies.
(Source:slackhq.com, Slack Prospectus)
As a workplace collaboration platform, Slack is known for its innovative user experience. It cleverly integrates instant messaging (IM) with various third-party applications, allowing its users to have all their tools – including file-sharing, chat, and project management tools – all on one platform, similar to 钉钉、企业微信 and飞书 in China.
In the relatively slow growth market of SaaS product, Slack’s ability to maintain its rapid growth extremely impressive. However, just as we were about to dive deep into Slack’s growth strategies, we noticed something strange: Slack’s web traffic actually barely improved in 2018. How could it be that a with such rapid growth could have such lackluster web traffic? What’s the catch?
Is Slack really the perfect story of growth? We decided to dig deeper by starting with Slack’s web traffic data.
How do we interpret Slack’s growth trends using web traffic data?
Before we dive into the details, we need to know why web traffic data is important to Slack. After all, as a workplace collaboration platform, surely the desktop application is more popular among users. So why are we looking at the web traffic?
First of all, unlike most consumer facing applications, Slack doesn’t automatically prompt users to download their app. As an enterprise facing tool, the web application is the initial starting point. Regardless of how they sign up, new users must first register on the website in order to use or download the application.
Meanwhile, registered users still have to use the website to access other functions, including third-party applications and upgrading their account to paid users.
We can see from Slack’s prospectus, where it claims that is uses a “web-based, self-service go-to-market approach”, that its website plays an important role in its marketing strategies. Simply put, Slack first attracts free users to its website with some marketing. Those users, after realizing Slack’s value, then recommend it to their teams. Slack then uses direct sales efforts to acquire larger organizations that already have Slack users as paid customers.
(Source: Slack Prospectus)
Combining the abovementioned Slack product design with its go-to-market strategy, the user journey is as follows:
- First, new users learn about Slack as a result of a range of marketing strategies and access its website and begin using Slack as free users.
- As the users enjoy their experience, they spread Slack to their team and organizations through word-of-mouth and invitations.
- As their organization become accustomed to Slack and begins to need paid features, they upgrade their account on Slack’s website.
- Slack’s direct sales team acts as a complement to the abovementioned “self-service” model and helps convert free users into paid users through seminars, executive activities, etc.
Hence, the website is essential to Slack’s key user actions – registration, download, and payment all happen on the website. These actions can all be tracked and quantified with website traffic data.
As we looked at Slack’s website traffic along its user journey, we noticed that traffic changes on key webpages show a downward trend.
(User journey of Slack)
1. Traffic growth on key webpages is slowing, new attempts made in user acquisition
Data shows that the proportion of web traffic on key pages didn’t increase; instead, it decreased slightly over time:
- The page after signing up (slack.com/get-started)
- the page for creating an organization (slack.com/create)
- the download page for Windows clients (slack.com/downloads/windows)
- the page for choosing your organization (slack.com/you-workspaces)
This implies that the percentage of users accessing slack.com that eventually sign up is declining, reflecting a lack of momentum for growing its user base.
Faced with this issue, Slack tried a new strategy with display ads. We analyzed the display ads network and found that the ads network through which it was advertised changed significantly over time. The number of advertisements on Google Display Network was gradually decreasing, while the number on Carbon Ads saw an opposite trend.
Google Display Network doesn’t target specific groups. Carbon Ads, however, has a target group – developers and designers. So, apart from Slack’s typical users (developers), is Slack trying to attract designers as well?
By analyzing the sites where Slack advertised display ads on, the proportion advertised on pixabay.com, a free image website, has been increasing, while those on sites with general user profiles has been declining, examples include Q&A community quora.com and RSS aggregation site feedly.com,
In summary, user acquisition through Google Display Network is declining year by year. Faced with slower user growth, Slack began to target more specific user groups, trying to find a breakthrough in user acquisition.
2. Free-to-paid conversion rate is on a downward trend
Slack can raise revenue by converting free users to paid users. One way to do this is through self-service. Slack offers three paid plans (Standard, Plus, and Enterprise) where users can easily upgrade via the website, except for the Enterprise edition which requires contacting the sales team.
However, web traffic data on the payment page shows that fewer and fewer users are willing to pay. The proportion of web traffic on the payment page and the number of people visiting these pages are both declining.
Public data also shows the drop in the paid user conversion rate. Figures from Slack’s prospectus show that 10% of FY2018’s profits came from free users, while in FY 2019, it dropped to only 8%.
Overall, Slack’s paid user conversion rate is decreasing.
3. The influx of non-targeted users generates new ways of using Slack
Slack obviously wants to attract more enterprise users as converting them would be more efficient. But through our research, we found that websites are increasingly beginning to use Slack as a forum for their fans, leading to more and more new non-targeted users join Slack.
Users are increasingly drawn to Slack’s excellent instant messaging – there’s no limit on the number of participants or Channels (i.e. chat groups). Messages can be synced on all platform simultaneously, and users can quickly join discussions with a click of the shared link.
(Examples of shared links)
How can we track this kind of usage? We can answer this question by tracking traffic data of Slack’s referral links. Referrals occur when a user signs up via clicking on an invitation link.
By sorting the sites that direct users to Slack, we find that Computers Electronics and Technology sites are the main traffic drivers. This is very much in line with Slack’s primary target users.
What’s more interesting is that two other types of sites that use Slack as a community: adult sites and cryptocurrency sites.
In the 2nd and 3rd quarters of 2018, adult site referral traffic rose significantly. Those websites created a variety of discussion groups. However, these groups all disappeared within a few months. It is speculated that Slack might have intervened since, after all, these could be detrimental to its brand as an enterprise application.
（Source: SimilarWeb, referral traffic proportions）
From the desktop referral traffic data, we found that in early 2018, more than 20% referrals came from financial sites.
Of these financial websites, the word ‘coin’ frequently appears in these websites’ domains. These domains direct to various cryptocurrency websites.
By comparing the trend of referral traffic in these websites to the trend of keyword “bitcoin” in Google Trends, we found a match. They were both rising in early 2018 before they started to decline. It seems that investors in the cryptocurrency market were also using Slack as a community.
(Keyword “bitcoin” Google Trends trend)
How can we quantify the trend of using Slack as a community? As mentioned earlier, to join a community, an invitation link to group needs to be generated. Hence, the times at which those invitation links appear reflect the trend of using Slack as a community. By plotting the monthly newly generated invitation links, we find that this trend is clearly growing.
On one hand, the influx of non-target users does not contribute to Slack’s paid user conversion rate. A high percentage of free users naturally occupies much of Slack server resources and bandwidth, costing Slack at least $6.67 per user per month.
On the other hand, more users are help Slack’s network effect, as they imply higher product value.
4. Product-driven to sales-driven transformation, relying more on Key Customers
By looking at Slack’s investment in product development and its job listings, we can see where its growth drivers are.
Slack has a great product experience, and it has a lot to do with its huge investment in the product at an early stage. Around July 2016, Slack’s hiring efforts were mainly focused on software engineers – SWE job listings were double that of sales. (But the media claim that Slack reached a $4 billion valuation company “without any salespeople” is wildly exaggerated)
(Job listing for July 2016, Source: Thinknum)
However, since February 2017, job demand for customer success teams and marketing teams have rocketed to the top 2, far outpacing the need for software engineers.
(Job listing for February 2017. Source: Thinknum)
This trend held until May 2019, where the top five job demand in hiring were mainly customer success, enterprise, sales, and marketing-related positions.
(Job listing for May 2019. Source: Thinknum)
Thus, product development is no longer Slack’s major focus; in contrast, sales and customer success becomes vital.
There are also clues to these trends in Slack’s financial reports.
Slack’s research and development expenses grew by 11.5% between the fiscal year 2018 and 2019, while sales and marketing expenses grew by 66.3%. Research and marketing accounted for a fairly share of the overall operating costs in FY2018, but by FY2019, marketing costs have reached about 46%. Therefore, marketing has become Slack’s focus.
(Source: Slack Prospectus)
What was the effect of Slack’s strategy shift from product-driven to sales-driven?
According to its financial report, in FY 2019, 575 key accounts of Slack’s 88,000 paid customers contributed more than $100,000, in other words, 0.6% key accounts generated 40% of annual revenue. In the two previous fiscal years, key accounts contributed only 22% and 32% of annual revenue.
Slack’s sales-driven strategy led to a revenue model that relies on key accounts – a significant difference from the initial revenue model that targeted small and medium-sized enterprises. Despite the benefit that relying on the key accounts generates revenue efficiently, the loss of one such account is also a much larger loss.
Overall, Slack’s strategy has shifted from product-driven to sales-driven. It seems that although “good product” might be a key factor in Slack’s early outbreak, it’s not enough to drive growth in revenue. A well-established sales model and after-sales service are more significant now.
5. Active key accounts’ retention rate is stable
As we mentioned earlier, key accounts play an important role in Slack’s growth. Hence, the retention of them should reflect if Slack’s business is table.
Through web traffic analysis, we obtain a monthly list of more than 500 active key accounts of Slack. Internet companies and software companies accounted for the majority of the top 10 users of non-enterprise (including Free, Standard and Plus plans) and enterprise plans. (One particular case, VMware, held a non-enterprise workspace in addition to the enterprise one. It is presumed that they did so to avoid expensive cost due to a large number of employees.)
Based on the monthly list of active key accounts, we can calculate their retention rate through cohort analysis. It should be noted that the data here are only the highly ranked users based on tracking their Slack’s workspace traffic. This calculation refers to the retention of active KAs.
(Active key accounts cohort analysis)
In terms of the monthly retention rate of Slack’s active KAs, its overall trend is stable despite fluctuations.
Thus, Slack’s key accounts are still relatively loyal, which implies the good performance of their customer success team.
6. Slowing growth in the integration of third-party applications
Slack is famous for its ability to integrate various third-party applications, such as GitHub and Google Drive. It also allows the development of new applications based on the open, documented APIs and developer tools. Described in the prospectus as The Slack Platform, this feature reflects its ambition to become an all-in-one office platform.
Slack’s growth has also encountered a few setbacks based on the traffic of app dictionary pages. Users who want to access third-party applications need to go to slack.com/apps, where we analyzed the change in traffic. We found that there were significant visits in May and June 2018 but there has been little growth since then. Furthermore, there has been a slight downward trend since 2019. It can be interpreted that fewer users go to this page looking for a third-party app.
Slack is now connected to over 1500 third-party applications, but this number was gradually built up in several years. Newly integrated applications will naturally attract users to use new features. If we can track the number of applications that integrate with Slack over time, we can see the pace Slack integrating new applications.
We found that when Slack adds an integrated application, it creates a dedicated page in App Dictionary on its website to introduce it. The introduction page sets up an outgoing link directed to the app’s official website. By tracking the changes in the number of outgoing links added, we can observe the changes in the number of newly added applications.
(Example of an external link)
Plotting the time curve when outgoing links were observed for the first time, we find that the number of new applications increases on the long run, peaking in the fourth quarter of 2018 but started to decline after entering 2019.
Thus, with the continuous integration of vast applications, their numbers are becoming saturated, and users are less interested in visiting the App Dictionary page.
7. Social media struggles to be a long-term driver for growth
Slack’s excellent user experience did attract a lot of users in the early days, and the word-of-mouth effect could be greater on social media. CEO Butterfield once said, “We bet heavily on Twitter. Even if someone is incredibly enthusiastic about a product, literal word-of-mouth will only get to a handful of people – but if someone tweets about us, it can be seen by hundreds, even thousands.”
In response, Slack offered users a place on Twitter to express their love for Slack. Initially, the main account, @Slackhq, created two “wall of love”s dedicated to posting users’ positive reviews of Slack (those sectors has been out of service since 2015). Since 2015, Slack sets up the account @SlackLoveTweets and transferred the “wall of love” service to it.
However, based on data from Social Blade, there were only 100 new followers of this account in the past 1 year.
(@SlackLoveTweets number of followers. Source: Social Blade)
And there is a downward trend of the number of tweets with negative monthly tweets posted, indicating that some users have deleted their tweets.
(@SlackLoveTweets total and monthly tweets. Source: Social Blade)
For now, Slack’s social media promotion has been less and less potent that it is hard to support Slack’s growth.
8. Rising threats from competitors
While Slack is soaring, its competitors built their rockets as well.
In a survey of 901 companies by Spiceworks, 21 percent of the respondents used Microsoft’s Teams in 2018, and only 15 percent used Slack. Compared to the results in 2016, Teams has become an ineluctable force.
Just as we were studying Slack, Microsoft released daily active user data for Teams in early July, and its DAU has reached 13 million. Looking back at Slack’s 10 million DAU data released in January 2019, Teams might have surpassed Slack’s if the DAU of the latter hadn’t improved much.
In addition to the challenge from Tech giant Microsoft, open-source enterprise collaboration tools are rising as well. Using a tool like Slack is, after all, risking a private data breach. Coupled with the high cost of subscriptions, the demand for open source collaboration tools rises. Open source can help enterprises save money, while private deployment ensures that data is well secured. Rocket Chat, for example, has been very popular recently.
Another competitor, Mattermost, aggressively placed a contrast between Slack and itself on its homepage to stress its advantages.
Discord, a product that looks similar to Slack, emphasized its voice-chat feature, and gained astonishing growth in users from the gaming market.
With a limited market size, the growth of its competitors indirectly suppresses Slack.
(Team collaboration software market capacity, source: Statista)
Overall, Slack is an excellent product, but it is facing obstacles – moderate growth in key webpage traffic, lower free-to-paid conversion rate, limitations in integrated applications, weak social media promotion, and threats from competitors all limit its room for growth.
All these problems can be clearly reflected in the stock markets: after the surge on the first day, its stock price fell for a month in a row, and now it is almost below its initial offering price.
But the stock price of Zoom, another SaaS benchmark listed in the same quarter, has tripled despite volatility and continues to rise.
Current investors, it seems, lack confidence in Slack’s growth momentum.
Slack, on the other hand, certainly doesn’t sit back and watch. There are signs that it is making attempts – optimization of user acquisition, the shift to the sales-driven pattern, the focus on Key Accounts, and so on.
From a financial perspective, Slack is still showing a healthy state of growth. Estimating according to financial data published in its prospectus, Customer Acquisition Cost(CAC) of Slack has been in fluctuation in the last three financial years, while the Average Revenue Per User (ARPU) has gradually risen.
If we look at the ARPU of Key Accounts who pay more than $100,000 a year, the indicator has risen to about 1.6 times in three years, showing a growing trend.
Slack’s Net Dollar Rate Retention was at 171%, 152%, and 143% respectively over the past three years. Net Dollar Retention Rate has been mentioned many times in Slack’s prospectus, reflecting its existing customer loyalty and ongoing ability to pay. Generally speaking, a NDRR above 100% is a sign of healthy development while if it below 75%, the company should re-examine its business model to see what went wrong.
Slack’s NDRR also ranked high in the industry, according to SaaS companies’ Net Dollar Retention Rate gathered by CrunchBase.
However, Slack’s Net Dollar Retention Rate is on the decline year-on-year. If this trend continues, Slack will be ripped of the advantages it has.
Competition in the enterprise collaboration software market is extremely fierce, and Slack, as a pioneer in this industry, must have accumulated rich experience along with its development. Especially in the current market, B2B growth hackers should continue to follow and learn from how Slack deals with obstacles and finds new space for growth.