Want to see what we’ve been up to since May 2020? Hint: We’ve blown up account identification using IP and cookie data. Read my recent post: In This New Era, Demandbase Unlocks Improved Target Account Identification with More IP and Cookie Data
Updated: May 19, 2020. See the new Account Identification and Advertising Data sections at the bottom.
Due to recent worldwide events, there’s been a surge of employees working from home. This presents a challenge for B2B marketers on several fronts. Among many things, events are being canceled and innovation becomes a new challenge, as discussed by my colleague Nani Shaffer last month. There’s also a new reality B2B marketers are facing today, which is explored in depth from our CMO, Peter Isaacson. Account-Based Marketers, and the vendors that support them, require adapting to these changing environments. One of the main challenges coming out of our current situation is how to continue to identify and market to companies when we are all working from home. Demandbase is well suited to help our customers with these new realities.
For those not aware, one of the elements of today’s Account-Based Marketing landscape is the ability to de-anonymize website traffic by looking at and identifying the visitor’s IP address. Some vendors (not Demandbase) power their “account identification” by solely relying on Reverse IP data. Reverse IP databases, analogous to the “Yellow Pages” of years past, are rarely updated databases that include the “domain name” associated with the IP address. While Reverse IP wasn’t designed to identify company visitors, it is accurate about 5-20% of the time when employees are on their corporate networks. In defense of vendors that use reverse IP, it’s easy to productize and 20% identification is better than 0% identification. However, when people work from home, this 20% identification decreases to 0% – since home IPs are associated with ISPs such as Comcast or AT&T. And here lies the problem. If employees are working from home, how can B2B companies find them online and market to them?
Demandbase has always taken a different approach with Account Identification – An approach that powers most of our solutions. At any given time, Demandbase has around 3.7 billion IP addresses that are mapped to a company or otherwise classified. We source our IP data from proprietary methods and use numerous 3rd party sources to validate & enhance our matching capabilities. Additionally, Demandbase has hundreds of millions of active cookies that supplement the identification.
When a visitor comes to our customer’s website, Demandbase first looks for our cookie ID. If present, we identify the company from this cookie alone. If there’s no cookie on their browser, we then look at the IP address to identify the company. Within the last couple of weeks, as employees work from home more, Demandbase has seen a slight increase in identification (especially with cookie-based identification) while our competitors are seeing a decrease. This is the result of not only our cookie database, but also by artificial intelligence.
To illustrate further, let’s use Microsoft as an example. As Microsoft employees work from home, Demandbase is seeing an increase in cookie-level identification. When cookie identification occurs, our AI engine still records the IP address of their network. Over time, if the same IP address is used by the same Microsoft cookie, the IP address would eventually be officially mapped to Microsoft.com, increasing our identification ability.
This IP to company mapping is accelerated if other unique cookies from Microsoft are using the same IP address over a short time span. The mapping of the IP is decelerated if we see other companies (based on cookie identification) using the same IP address – Our AI model will think the IP address is a coffee shop, an airline club, or any other public wi-fi network. We, therefore, wouldn’t map the IP to a company to minimize false positives that occur on account identification.
The constant assessment our AI makes with matching IPs & cookies to accounts is what allows our customers to see continued success. Unfortunately for other vendors, who haven’t had the bandwidth & resources to build out such a sophisticated data set with constant AI evaluation, they’re being highly affected by recent work-from-home orders.
Note: Demandbase will always prefer false negatives when it comes to account identification as opposed to false positives. False positives will give the impression of higher match rates at the cost of accuracy. Any ABM vendor who proactively advocates doing “match tests” between competitors will always have a significant amount of inaccurate identifications. Taking marketing & sales actions based on incorrect data is completely counterintuitive to an effective Account-Based Marketing strategy.
Demandbase is seeing an increase in intent behavior collection since we’re now associating more IP addresses with companies as employees browse the internet on their home network. The intent gathering methodology we deploy is not impacted. Other intent vendors are seeing a drop in intent collection because they rely on 3rd party data sets that they purchase and have no control over.
For context, Demandbase leverages artificial intelligence & natural language processing by taking opted-in visitors from across the internet & analyzing the terms & phrases that those business readers are engaging with. The power behind this was outlined above – starting with our ability to identify accounts based on their IP address and/or cookies.
Once we confirm which company is reading the article, we then filter out the noise on business content pages & narrow in on actual product research. Unlike Demandbase, our competitors don’t own or maintain their own datasets – they either partner with 3rd party intent vendors, who categorize internet behaviors, or they gather intent data from Meta Tags within website HTML – which wasn’t at all designed to be indicative of buying intent. It’s more relevant for SEO (and sometimes reminiscent of the keywords added to the bottom of a Craigslist ad to pull in viewers for an irrelevant product). I can’t stress the importance of accurate & scalable intent data since it’s a big predictive element when determining what ABM program or stage a company belongs in.
Demandbase incorporates the aforementioned topics to power our advertising platform. Demandbase Advertising is unique because we have the ability to prioritize impressions to the buying committee at your target accounts. We do this not by persona or title-based targeting, which can be unreliable – but rather based on real-time intent behaviors within a company.
For example, Let’s say your company sells Marketing software & you were tasked with reaching your target accounts. Since you would want to advertise to the Marketing Department within the companies, you would ask Demandbase to listen for intent behaviors within the accounts related to “Account-Based Marketing,” “Lead Generation,” or “Best ABM Vendors,” etc. The employees researching or reading these relevant articles will undoubtedly be the decision-makers, influencers, and the entire buying committee.
Demandbase will then serve impressions to the “high-intent” users within your target accounts. Because we’ve established in the first two subjects of this blog that Demandbase’s account identification & intent behaviors are increasing, our ability to advertise is more efficient as people work from home, especially within countries that inherently have more cookies present on web browsers.
The above graph represents impressions Demandbase has access to and the identification method. By mid-February, we started to notice an increase in cookie-based advertising as employees started to work from home. At first, employees worked from home by choice, followed by directives within company management. By March, we’ve seen a steady increase in the bifurcation of cookie-based advertising as opposed to IP-based advertising. For the time being and in the immediate future, this allows for much more targeted advertising with Demandbase because historical intent behaviors within cookies are tracked by Demandbase. Our smart bidder will dynamically bid higher for employees that have a lot of historical intent behaviors as opposed to cookies that have a lesser amount of intent (or IP-based identification that has a single relevant content consumption behavior).
Since our solutions were built on the above technologies, all of the features our customers have grown to love & find value in are still fully viable. This includes our predictive abilities through account selection, web analytics integrations, ROI reporting metrics through our ABM Analytics, Site Analytics, and campaign performance tools, our ability to personalize web experiences, shorten & enrich forms, and perhaps most importantly, our ability to power sales with real-time data & alerts. In the new marketing & sales reality today, all of Demandbase should be leveraged to keep pipelines healthy & forecasts on-track for success.
As current events unfold, rest assured Demandbase will be closely monitoring our data and will be constantly communicating best practices and recommendations to our customers. For now, we are recommending to focus on all account-based digital efforts–especially since marketing and sales KPIs are harder to achieve with the more traditional marketing initiatives and channels.
The graph above shows Demandbase Match Rates on global pages and separately within our customer base. The small dip in mid-March is what prompted our AI engine to adjust to the major economic shift to WFH. Despite the increase in consumer traffic across the web (as indicated in blue), Demandbase has slightly increased our company identification rates (+1-2%) within our customer’s websites.
Note: This graph shows the % of campaigns that are on-pace with delivering the full amount of impressions promised over the flight dates of ad campaigns. WFH is not negatively affecting our advertising product. In fact, quite the opposite as the next graph shows.
This final graph represents all of the campaigns Demandbase is running on behalf of our customers. We’re proud to share that Demandbase has seen a significant increase in % of campaigns that are well above a .04 CTR. Keep in mind that Demandbase is serving ads to the buying committee based on their real-time intent behaviors.
What better traffic is there?