Case Studies
Development In A Data Void
Making A New Audience For A Classic Product
Context — A Product To Help Small Businesses Flourish Locally
Recruiting and managing a delivery staff can be a challenge for smaller brands. It involves significant time and financial investments, from finding and training personnel to maintaining their vehicles. And poor delivery experiences, like damaged food or long wait times, can adversely impact a merchant’s ratings and local reputation.
In response to this common growth hurdle for smaller businesses, DoorDash offers a product called Drive (later rebranded to Drive On-Demand). This solution gives merchants access to DoorDash’s extensive network of delivery personnel and infrastructure, which enables them to make deliveries in their neighborhood safely and reliably without having to worry about the overhead of managing their own fleet of drivers and cyclists.
When businesses leverage Drive, they can shift their focus to what truly matters: providing customers with exceptional, memorable experiences that keep them coming back for more and encourage them to spread the word about a local gem through positive reviews and word of mouth.
But somehow, after six years of life in the market, this product had never before been the focus of a dedicated email marketing campaign. Our recently established team, responsible for marketing automation and email-centric strategies using HubSpot, was therefore assigned the task of launching its inaugural nurture series.
This nurture campaign was designed with four key objectives in mind: to amplify brand awareness for an exceptionally valuable product; to initiate meaningful conversations with relevant prospect; to drive product sales and revenue; and to strategically position ourselves for ongoing outreach, setting a foundation for sustained growth in promoting the product.
Challenge — A Void Of Usable Data
In standard practice, our team relied on both custom and out-of-the-box data fields, populated through manual HubSpot imports or automatic transfers from Salesforce, to define the target audience for various campaigns, including those for email marketing. Within our extensive database of 2.5 million prospects and customers, we’d use a mix of these fields to engineer just the right segment.
However, in this case, a significant amount of data was unreliable, partial or downright missing. We realized this data void was primarily due to the historical lack of proper database maintenance. Before our team was formed, our HubSpot instance had accumulated hundreds of thousands of disparate contact records from various sources, events, ads, regions, verticals and more. Unfortunately, these records had not been cleaned, validated, standardized, deduplicated or kept up-to-date—which rendered them all but unusable.
Maintaining good data hygiene is crucial for organizations to make informed decisions, generate accurate reports and ensure the effectiveness of various data-driven processes. The repercussions of this issue were, for us, most apparent in the segments of the database relevant to our marketing campaign for Drive. Our goal was to create a segment of U.S.-based merchants in delivery-friendly industries who expressed interest in the Drive product and were not actively engaged by a salesperson. To scrape together a segment that fit these parameters, we would have to overcome this bad case of data mismanagement.
The challenge became even more pronounced when considering how to plan essential campaign elements—messaging, cadence, tests—without certainty in our ability to build a well-targeted audience. For example: Should we craft generic emails to cater to a broad audience, or trust in our competency to identify actual business owners in a complex database?
Given DoorDash is a three-sided marketplace, anyone in our instance could be a merchant, customer or dasher. An incorrect approach could lead to adverse outcomes, such as an influx of spam reports and unsubscribes, affecting overall channel performance and triggering concerns from our HubSpot account manager.
Resolution — The HubSpot Embed Code
Having to deviate from our usual methodology, we constructed the target audience for Drive’s inaugural email marketing campaign by delving into data that the marketing team had never previously examined for campaign development.
Our initial step involved cataloging all conceivable Drive-related landing pages, spanning from the core product page to specific case studies, blog articles, ebooks and webinars. Over the years, any web property integrated with HubSpot’s embed code had been transmitting behavioral data back to our system, recording site visits for individuals with known email addresses.
Subsequently, we implemented automations to identify individuals in HubSpot whose contact records contained Drive-related values in web fields, such as First Page Seen and Last Page Seen—basic yet potentially robust options readily available.
While this approach had its limitations (“Hey, where is Middle Page Seen?!”—this somewhat nonsensical field had no analog in HubSpot), it served as a reasonable starting point. Given Drive’s modest organic promotion over its six-year existence, it became noteworthy if someone had visited a Drive-related webpage. Such visits could signify a merchant actively seeking a solution to their delivery challenges.
Specifics — Two More Data Layers
But we aimed to go to market with even more confidence in our audience build. We knew what we didn’t know—the known unknowns had precipitated the entire web-focused part of the solution—but we agonized over what we potentially didn’t know we didn’t know.
The unknown unknowns.
For example, it wouldn’t be inconceivable for someone who was not a merchant, but rather a gig worker interested in dashing or making food deliveries, to head to Google and search for terms like “DoorDash” and “drive.” That person, technically, would have also been a visitor to one of the Drive pages we had cataloged—logging them in our automation as an audience member for a B2B campaign not intended for them.
To strengthen our targeting, we therefore added two more layers to the segment, refining our scope to hone in on contacts who were also:
Classified as recent B2B engagers, meaning they had accessed a webinar, completed a form, opened or clicked a marketing email or downloaded a report sometime in the last 30 days;
And positioned at the bottom of the marketing funnel, i.e., already owning traits or having taken actions that indicated they were ready to convert.
By incorporating these two additional layers—prospects positioned closest to the edge, coupled with the likelihood of inertia nudging them toward conversion—we figured we maximized our odds of executing a successful campaign.
Our team then programmed to send messages on the best days and at the best times for merchants, trusting our approach to the data would indeed let us reach this audience. We balanced this, though, with purposeful broadness, touting value propositions that would appeal to business owners in all of the verticals serviced by the DoorDash app: no need to worry about training delivery-people, tracking them, managing a fleet, or buying and repairing vehicles.
Further, we fleshed out the series with bottom-funnel emails. Some reassured prospects with success stories and testimonials; some created a sense of urgency by highlighting our the time-sensitive nature of a discount or special offer for merchants who partnered with us that day; and some preemptively answered finer-point questions, such as those around pricing tiers and integration with certain point-of-sale systems.
Results — A Chain, A Mom & Pop Shop, And A Receptive Audience
During its first month of life, the first-ever email nurture for Drive resulted in the successful acquisition of two new partnerships for DoorDash.
Wingstop, the third-fastest-growing American restaurant chain,¹ as well as Nectarine Grove, an organic bakehouse in California, both chose to collaborate with DoorDash and sign up for Drive. These partnerships were and are valuable additions to the ever-expanding network of merchants DoorDash supports—and is supported by—through its core app.
Channel outcomes were almost as positive as business ones: Read rate was 68.9% (115% higher than global benchmark)² and unsubscribe rate was 0.09% (82% healthier than best practice).³ To us, this indicated our resourceful approach—filling in the data void when and where we could, pulling together, bit by bit, a reasonably well-shaped target audience—had worked.
High marks for read rate (which measures how many people read an email from top to bottom) and a very low number of unsubscribes meant this campaign had likely earned the interest and trust of the segment we reached out to, laying the groundwork for a long-term relationship that could lead to conversions down the road for those who hadn’t converted this time and upsells for those who had.
We would continue to architect more campaigns for this audience, which over time would give us the core insights we’d been missing at the start—lifting overall database hygiene and enabling us, ultimately, to flesh out an even fuller picture of these prospects.
_____
¹ Between 2014 and 2016, Wingstop was the third-fastest-growing restaurant chain in the US as measured by both system-wide sales and unit growth, according to Nation's Restaurant News (archived links here and here).
² Industry-agnostic standard provided by Campaign Monitor. Outperformance based on average read rate across all campaign emails.
³ The best-practice standard (≤0.5%) is echoed across countless reputable sources, including Shopify, ActiveCampaign and Bluecore.