Growth case study: B2B growth
A detailed look at the journey to spin up Pinterest's B2B growth effort
A lot of literature on growth goes through a specific project or concept. This post is a bit different since it is composed as a case study. I chose this format since I often want to understand all of the context that led to a project being successful. I am hopeful this step by step case study can give the reader the necessary background to understand how the pieces fell into place. Thanks again for reading!
Introduction
Going into 2020 Pinterest wanted to focus more on short and medium term revenue growth (6-12 month revenue initiatives). The SMB Growth team at the time was focused on the long term strategy (2-3 year time frame) of building the self-serve ads business. This ended up with the team pivoting to accommodate the new goal. We decided to focus on the sales funnel and began specializing in B2B growth to drive revenue more immediately. This initiative ended up leading to new channels that represented 40% of opportunities for the SMB sales team. This post breaks down each stage of the initiative and summarizes the learnings and outcomes from each.
0-2 months - Pre-seed capital
Before a single line of code was written the idea needed to be fleshed out. Akin to the pre-seed stage there wasn’t any software built yet. The leadership team had a couple months to create a strategy, develop a topline metric and align with stakeholders.
Team
1 Engineer
0-1 Product Manager
0-1 Engineering Manager
Context
The general idea was to improve the sales funnel using engineering, product and design (EPD). Pinterest had focused on this in the past, but had begun focusing only on the largest advertisers leaving the sales teams focused on SMBs (small and medium businesses) with less EPD support. Focusing on SMBs aligned well with how a growth team operates given experimentation, measurement and automation are much more valuable at scale. For example if the total population of customers is in the hundreds it is much more efficient to just have a sales rep contact each one, while if it’s in the thousands or hundreds of thousands automation becomes a competitive advantage.
With this in mind the team took a look at the sales process/funnel and began to develop the strategic direction.
Inbound SMB sales process
Strategy
Although the graphic above seems simplistic each step has dozens if not more actions required to move to the next phase. In order to keep the team focused we decided to hone in on generating new leads and determining if they are qualified. User problems in this area would be primarily solving problems for SDRs (sales development reps) and help meet new customer revenue goals.
The three strategic directions decided upon were: lead generation, sales outreach and signals/scoring. Each area was backed with data and qualitative feedback from our sales team.
Lead generation
Given the large footprint of Pinterest’s creator base there was substantial potential for turning existing businesses on the platform into advertisers. There was also a large opportunity in optimizing inbound lead generation given only about 20% of businesses were being even considered to speak with a sales representative.
Sales outreach
The largest population of leads that were qualified but didn’t convert were those who were unresponsive to outreach. Many of these businesses even raised their hand to talk to a specialist demonstrating their high intent. With a focused effort new outreach methods could be built and a shorter turnaround time between inbound interest and outreach seemed feasible.
Signals and scoring
In order to accurately determine whether or not a lead is qualified a significant amount of data is collected beforehand. Most of this information was manually collected by an SDR or was partially filled in by an enrichment provider. By automating data collection to increase coverage and using machine learning to determine qualification a substantial improvement in efficiency was possible.
Outcome
Once the initial strategy was in hand the team needed to align with sales to officially kick off the project. The primary way this was accomplished was by aligning EPD topline metrics with sales goals. Due to a longer sales cycle opportunities created ended up being how the team was measured. Leadership also agreed there would be a revenue look back once more data came in to understand actual business impact.
2-4 months - Seed capital
With a strategy in hand and topline goals set the team needed to make resource allocation decisions. The team was already fully staffed so we were deciding between a full or partial pivot. We ended up doing a partial pivot because we didn’t want to see the self-serve efforts completely neglected. This decision also made the team more nimble since it’s easier to lead 3-4 engineers then 7.
Team
3-4 Engineers
0-1 Designer
0-1 Product Manager
0-1 Engineering Manager
0-1 Product analyst
Strategy
Product
The three high level strategies laid out in the pre-seed stage needed to be sequenced. This was primarily due to SDR capacity constraints. For example doubling the number of leads sent to the sales team would only result in an increase in the time it takes for leads to speak with an SDR. Which studies have shown lowers the overall conversion rate of leads becoming paying customers. So in order to drive net new revenue we needed to either free up SDR capacity or increase SDR efficiency. Setting up a solid foundation was the top priority even if everything we built did not directly impact revenue.
Data
Often in the early stages of an initiative visibility is the first priority, but in this case the sales team had sufficient monitoring available. This allowed the data team to focus solely on making the product better by focusing on the signals and scoring strategy.
Tactics
As mentioned before setting up a foundation to drive revenue was the top priority. We decided our initial focus would be on freeing up SDR capacity since it was easy to measure and had high impact. In parallel we began building out different ways to drive net new leads through offering more entry points for a user to reach out to sales (inbound lead generation).
(Threshold represents lead score) Lead scoring had the potential to significantly reduce lead count with a minimal amount of revenue lost.
Opening up the contact sales form to more businesses during the sign up process could essentially double lead generation
Outcome
We drove a massive reduction in the number of low quality leads through lead scoring. Specifically we doubled the chance a lead would end up booking a meeting with our sales team, in other words we doubled the capacity of each SDR.
As the capacity increased, sales leadership let us know SDRs were sitting idle, but that didn’t last for long as the new experiences to drive leads came online. By the end of this period our new lead generation efforts made up about 20% of businesses who directly reached out to sales. With everything else being equal this would translate into a greater than 20% increase in net new revenue. However we didn’t expect to see the rest of the funnel stay the same, since each additional advertiser is usually lower value. Even with this caveat we were excited to see such strong results.
4-8 months - Series A funding
After the initial success, we felt more comfortable investing everyone's time into this new initiative. The sales team also was focusing more on SDRs and hired a new leader to drive impact at this part of the funnel. At this point the team could be compared to a Series A startup that had shown some success but needed to show consistency in order to get staying power.
Team
4-7 Engineers
1-2 Designers
0-1 Design Lead
1 Product Manager
1 Engineering Manager
1 Product analyst
Strategy
Product
Seeing initial success on using app traffic to drive lead generation we wanted to double down on this approach. The main focus up until this point was on new users, but we had been building the infrastructure to allow us to target existing users which made the strategy viable for a wider audience. We also wanted to look into other channels of lead generation, specifically outbound (off platform), since this was also a large portion of net new revenue.
The sales outreach strategy also became a focus as we added additional engineers to focus on B2B growth. Although in order to utilize new outreach methods technical infrastructure and security reviews were needed. Even with a longer payback period we believed this strategy would pay off in the end.
Data
The major increase in sales efficiency, through increasing conversion and decreasing low quality lead volume, turned out to be a much bigger deal for the sales team then we initially expected. Even without a statistical view of revenue impact, sales leadership was already seeing the benefits of the work and asked for more improvements. This shifted the lead scoring strategy to focus on collecting more signals in the short term so we could drive large model improvements in the long term.
Tactics
Taking advantage of the new infrastructure to target existing users was our top priority given we had spent significant resources getting this propped up. Executing on this was simple since we only needed to employ basic conversion tactics. We settled on finding the most visited surfaces and then targeting the most likely to convert users with an out of the way ‘Contact sales’ button.
As mentioned above the lead scoring work had been well received so we intended to focus on laying the foundation for this effort to continue well into the future. One of the primary areas to collect data was during new user onboarding which had not been updated for years. We planned to completely redesign it to remove superfluous steps and add new steps that we had predetermined would increase the strength of our lead scoring model.
Coordinating closely with sales leadership we wanted to bring meeting scheduling in app. Up until this point leads would receive emails that asked for times or led to a scheduling portal. Following the theme of utilizing in-app traffic, we believed we could see an incremental increase in meetings scheduled with sales. The in-app scheduling component would only show up when the sales team proactively reached out to the business in question, thus creating a new channel of sales outreach if a business ever visited pinterest.com.
Allowing businesses to reach out to sales for help at any point in their journey
A redesigned onboarding to collect further signal from businesses
Scheduling in app to allow for leads to instantly book meetings
Outcome
By the end of this period we saw a large acceleration in all of our KPIs, the most important being a 2x increase in net new quarterly revenue year over year. This was partially due to the large impact we had in the past 4-5 months but it also had to do with the long term nature of our customer life. The work we successfully completed in the seed capital stage was beginning to pick up steam. For example the inbound lead generation work from this initiative began to compound. This led to about 40% of all businesses who came in through inbound sales channels to be from these efforts.
The sales outreach work also ended up having a step change impact. We saw a 30% increase in the number of meetings booked with our sales team due to this new component. Setting up the infrastructure, getting security clearance and coordinating with sales was a lengthy endeavor (4-5 months) but the impact ended up being worth it. This result also opened the door for more iterations and new audiences to capture.
Conclusion
The coordinated efforts of the product, engineering and sales team had a measurable impact on revenue and new business opportunities. The results are clear, but the short term benefits aren’t as interesting as the long term implications. If a small team can drive this much impact, what could doubling the team size do? How many more opportunities lie within this feature space? It is difficult to spin up an initiative from scratch, but like any investment there is potential for the benefits to compound. Potentially making the inherent risk in trying something new worth it.
These case studies are meant to explain some of the processes for building a successful growth initiative from scratch with the below being the primary takeaways.
0-2 months - Pre-seed capital:
Come up with a loose but well vetted strategy to focus the efforts of the team
Aligning topline goals with cross functional partners is mission critical
2-4 months - Seed capital:
Setting a foundation for sustainable growth is always a top a priority
A bigger team isn’t always better
4-8 months - Series A:
Large investments can drive a step change in KPIs
Foundational projects may be more impactful than what data suggests