Snowflake’s Marketing Intelligence team delivers near-real time analytics to many functions across our Marketing, Sales, and Finance teams, making Snowflake a truly insights-driven organization.
We achieve this feat through analytics dashboards that deliver insights into the effectiveness of the marketing campaigns we run. Marketing stakeholders know at a glance exactly what is happening with each campaign, empowering them to make decisions based on data-driven insights.
We have also built data science applications that forecast sales pipelines and predict the best accounts and leads for Sales to pursue through account scoring and lead scoring. Powered by the continuous influx of data, our self-learning machine learning (ML) models optimize these predictions every day, which improves outcomes and enables close alignment between Sales and Marketing.
For many organizations, this type of marketing analytics may sound difficult, if not impossible, to achieve. Perhaps data lives in too many locations and is challenging to aggregate, or maybe data remains siloed in the many applications and systems your Marketing team uses.
Regardless of the reason, marketing analytics teams should have a single source of data if they want to provide a comprehensive view into all marketing programs and initiatives. Near real-time analytics demands it.
The power of marketing analytics in action
Here are some examples of the analytics dashboards and data science applications we have built. All of them rely on data from many SaaS applications we use in Marketing at Snowflake. The key is to centralize all of that siloed data. It’s a taste of what you can achieve with marketing analytics when you centralize data into one data platform, ensuring all data is easily transformed and ready for analytics.
Campaign reporting
For our demand generation team, we provide dashboards to track campaign responses and advertising spend. This empowers the team with near real-time visibility into campaign dollars spent versus outcomes. Our campaign reporting dashboards show factors such as how much money was spent, how many responses were generated, and what the weekly and even daily trends are by advertising platform and UTM parameter. In addition, we built a campaign attribution model to understand the campaign engagement journey for each account, along with the impact on the sales funnel in order to measure ROI. These insights enable marketers to make fast, insights-driven decisions about which campaigns are the most effective, as well as to plan campaigns for the future.
Account based marketing
Demand generation segments large audiences and markets to them via various channels. Account based marketing (ABM) targets specific named accounts with highly personalized content. At Snowflake, we built analytics dashboards that provide a holistic view into all ABM campaigns. We include enriched account information from additional data sources, which allows our ABM marketers to personalize campaigns at an even deeper level.
On a side note, ABM providers deliver manual reports against data, which may lead some marketing stakeholders not to care if ABM data is siloed. This mindset represents a roadblock that should be removed. The only way for marketing analytics teams to deliver a comprehensive ABM view is to combine ABM data with other relevant data sources.
Field marketing performance
To support field performance and drive alignment between Sales and Marketing, we deliver personalized dashboards where each field marketer can see their territory, accounts, and what kind of campaign engagement occurred over the last day, week, month, quarter, or year.
We started with our marketing database, including everything from ABM target accounts and campaign engagements, to field campaign events and broader company-level campaigns, as well as pipeline health. We then segmented the data and aligned it with the top-to-bottom field hierarchy within sales. Because all data is lined up and categorized, we now power this shared view between Sales and Marketing, and have built a strong alignment to drive targeted account engagement.
Partner marketing
While analyzing the impact of partner marketing on campaigns can be difficult for many organizations, we have built dashboards that provide complete visibility into different partner types and programs.
Because we collect so many different measurements, there are a multitude of ways to look at the data in order to evaluate partners. These near real-time reporting capabilities enable constructive conversations with partners around expectations and incentives.
Some dashboards show how easy it is to view campaigns by event, see how many people responded, and line up campaigns with different sales territories. We can also examine how each partner helped to generate demand or penetrate different industries. These insights enable Snowflake’s partner marketing team to sort partners into different tiers, based on insights-driven factors such as involvement and engagement.
Lead scoring
Because we use a single source for data, we have built ML models that deliver predictive lead scoring for every lead we receive. We have a fully-automated and ML-powered engine that produces near real-time predictions via a daily training algorithm, based on all data flowing into these scoring applications.
Our model is also self-learning, which means a new model is generated every day. The algorithm today is different from yesterday because the model is constantly learning, which is the real power of ML. As a result, we achieve our business objective around identifying the highest-potential target leads for Sales and Marketing.
In addition, we help prioritize the outreach sequences for sales development reps (SDR) in order to improve their productivity. Since the most precious resource for SDRs is their time, lead scoring helps them focus on the exact leads that are most likely to convert. As a result, the SDR team is experiencing a 50% improvement in lead-to-meeting conversion—a percent that will only continue to increase as the model refines itself with additional data. In the longer term, we plan to push scoring data back to advertising platforms so they can improve targeting or retargeting, as well.
Pipeline forecasting
We built an ML model that takes the guesswork out of pipeline forecasting. Rather than run CRM reports and conduct ad hoc analyses, our model predicts how much pipeline will be closed each quarter and how much will move into future quarters.
This ML model is extremely complex and relies on historical pipeline data and hundreds of other data points and signals to understand whether a particular opportunity will be closed, and when. It also uses data at the sales rep level, including how much new pipeline will be created for each rep and how much will be closed in each quarter. This helps sales leaders understand where they need to hire more reps based on where opportunities exist.
Because sales and marketing stakeholders see what is predicted to move through the pipeline and close, as well as a clear pipeline view against booking targets, they can line everything up to deliver on their objectives and drive business outcomes.
Insights-driven decisions enable cross-team alignment
As these examples illustrate, marketing does not work in a vacuum. By maintaining and building on these ML models and analytics dashboards, the marketing intelligence team is providing a centralized source of near real-time insights which can, and should, be consumed by other teams to magnify their impact. Teams such as Sales and Finance are empowered to speak the same language as Marketing and hold the same underlying assumptions and pertinent information. As a result, it’s easier to drive alignment around shared and complementary goals.
Field performance is the perfect illustration of the alignment we bring to Sales and Marketing. By joining together and analyzing marketing campaign data and sales performance data, stakeholders on both teams are empowered to make better decisions through insights-driven conversations and shared goals.
Our pipeline forecasting is another example. The marketing intelligence team provides valuable information to Marketing and Sales on the rate at which we are filling and emptying the pipeline. But these insights also deliver value to the Financial Planning & Analysis (FP&A) team, which is responsible for forecasting and budgeting. That’s because revenue is impacted by how well the Sales team converts leads, which is impacted by how fast the marketing pipeline moves, how effective our marketing campaigns are, and how accurate we are with our lead scoring.
The same is true in reverse: If we demonstrate through campaign reporting analysis that Marketing is achieving high levels of success with certain campaigns, then the data supports a conversation with Finance where marketing stakeholders can request additional budget to accelerate, boost, or replicate these types of campaigns.
It all starts with the right technology
The fundamental problem facing marketing analytics teams today is figuring out how to bring all your data together. Without a unified and single copy of all your data, it’s extremely difficult to deliver robust analytics and build models that provide timely predictive and prescriptive insights.
A modern cloud data platform solves these challenges by delivering a central repository where all data resides in a single location. A platform built on a multi-cluster shared data architecture makes it possible to bring large volumes of varying data together quickly, combine data sets, run advanced analytics in near real-time, and build machine learning models.
To truly optimize your marketing programs and predict your ROI, your cloud data platform should:
- Scale the continuous ingestion of massive amounts of data from all your applications, and provide the compute power to run a near-infinite number of concurrent data workloads without impacting performance.
- Centralize and transform varying data—structured, semi-structured, and unstructured—that is produced and siloed in different SaaS and cloud data stores and push that data back to external applications.
- Securely share live data across your marketing organization with other stakeholder groups and with other business partners, while easily acquiring external data sets to further understand customers and inform your sales strategy.
- Enable modern data science to handle virtually any volume of data at speed to power business intelligence and advanced analytics, and train machine learning models.
- Provide modern security and governance features to protect data, meet industry and regional data regulations, and track where data lives, who is using it, and enable a single copy of your data that a near-unlimited number of concurrent users can rely on.
Building a diverse team with a thriving culture
Besides the right technology, we are building a diverse team that is globally distributed to drive the continuous impact for the present and the future. Marketing intelligence is a technical domain, but also needs to be creative and constantly innovate. Every member of the marketing intelligence team wears multiple hats, building best-in-class analytics dashboards, developing predictive data science applications, telling data stories creatively, and driving thought leadership activities to win customer minds and influence the industry.
Delivering insights today and in the future
We have a customer-first value at Snowflake that drives everything we do. For my marketing intelligence team, our customer varies. While it’s usually Marketing, Sales, and Finance teams, it can also be external customers. Recently, I spoke to a customer who didn’t have a lead scoring model in place and was looking for ideas on how to scale operations with data. These conversations are becoming more frequent and make me realize that Snowflake is poised to help in many ways.
Today, we are enabling Snowflake’s global marketing team to become the industry’s most insights-driven team. Perhaps tomorrow we will be executing against that goal for a much broader audience by offering our scalable analytics dashboards and ML models to customers. The beauty of data and machine learning is that we are always learning and improving, which means there’s no end to the insights and opportunities we can uncover.
The post Deliver the Future of Marketing Analytics Today appeared first on Snowflake.