Conquering big data
IMM looked to IBM Business Partner eCapital Advisors for recommendations on how it could harness new technology to gain the insights it needed.
Ken O’Brien comments: “To continuously improve the services we provide to our clients, we needed to make better use of data, especially in areas such as media investments. Initially, we considered building a reporting tool for ourselves using open-source solutions. But eCapital Advisors were able to demonstrate that IBM could provide everything we were looking for, so we could get started so much sooner. We decided to deploy a complete, integrated stack of IBM Analytics solutions that would empower us to develop a wide range of different analytical capabilities.”
IMM ingests data from its clients, combines it with data-sets from third parties such as ad servers, publishers and attribution vendors, and uses the IBM solutions to analyze it on a large scale and at high speed. This enables the company’s analysts and data scientists to extract rapid insights into impressions, click-through rates and conversions for online ads.
Ken O’Brien comments: “As far as we know, the analytics capabilities we now have at our fingertips are extremely rare in the advertising industry, giving us—and our clients—an edge over the competition.”
Sean Baker, Chief Client Officer, agrees: “On their own, our clients might not be able to justify investing in industry-leading analytics technologies. But because our agency can scale the tools across many clients and verticals, we can offer them the advantages of advanced analytics that would normally be out of reach.”
IMM’s big data analytics architecture is built around IBM® BigInsights® for Hadoop, IBM PureData® System for Analytics and IBM Streams, with IBM Cognos® Business Intelligence and IBM SPSS® Modeler providing reporting, data mining and predictive modeling capabilities. By opting for a five-year financing deal from IBM Global Financing, the company was able to take advantage of these solutions while minimizing the impact on its cash flow.
Ken O’Brien says: “The IBM PureData System at the core of the solution crunches through data incredibly quickly. It enables us to use Cognos to generate reports comparing campaign success against budget at much greater speed and accuracy than before. And with SPSS Modeler, we can begin to forecast ROI and pinpoint the best times to purchase media.”
Gina Lee continues: “With Cognos, we can now share significant stats such as cookie pool penetration of retargeting media, cross-device conversions by path, and influence on offline sales with online activations. With this daily feedback loop, we can continually hone and adjust campaigns for optimal results, and make better purchasing decisions and recommendations to clients.”
Since IMM works with clients across multiple industries, one of the key benefits of the new platform is that it can easily analyze a wide range of data-sets, empowering IMM’s data scientists to slice and dice the results and find the correlations that are most relevant to each specific company.
For example, IMM is putting its analytics solutions to work to support the media campaigns for a major telecommunications provider.
Ken O’Brien explains: “In the telecommunications industry, the principal factors that affect purchasing behavior outside of advertising include competitive offerings, consumer contract cycles, holiday drive periods, product availability, and offline and in-store experiences. With IBM Analytics solutions, we are able to incorporate insights from these macro trends into daily decision-making around how advertising dollars are spent.
“For example, analyzing store data can tell us how many people download or print out a coupon and redeem it in-store—so we can trace the entire customer journey from online ad to in-store purchase. And by combining GPS data on store locations with advertising data-feeds, we have been able to build a competitive map that shows which other companies are competing with our clients in a particular area. The competitive map can help answer questions such as: ‘do I need a stronger offer in areas where competitors are especially nearby?’ or ‘Is a competitor campaign impacting the performance of mine?’
“The IBM solutions also show us daily sales for each type of device, so we have a better handle on customer demand than ever before. We can use this insight to proactively suggest new offers, catalog placements and price points to the client.”
For a large restaurant client, IMM is exploring the relationship between weather data, time-of-day and door numbers to maximize the impact of its digital ads.
Ken O’Brien elaborates: “Weather is a hugely influential factor in whether people decide to go out to eat, so it makes sense to push more ads out when a fine evening is forecasted. For a restaurant, the type of ads aimed at families and those aimed at young professionals going out after work will be very different.
“Using the IBM Analytics solutions to analyze order information, we can pinpoint what time of day different groups tend to visit each restaurant. We can then focus on placing ads at the right time, on the right websites, and in the format most likely to appeal to particular customer segments. A voucher for free chips and dips could strike a chord with the after-work crowd, while a coupon for sundaes could draw in families. Now we can see almost immediately what the impressions, conversions and returns are like for each offer, and adapt the campaigns accordingly.”
Ken O’Brien continues: “We are already at a stage where many of our clients are amazed by the new insight we can now deliver—but we see this as just the start of our journey. We’re excited by the prospect of moving into cognitive analytics with IBM Watson™ Analytics.
“For example, at the moment we’re slicing and dicing all of our data sources and putting them in a format where our data scientists and our clients’ analysts can work with them and answer questions from the business. But if we hooked up Watson Analytics and pointed it at that same set of well-organized data, the business users could start answering their own questions.
“So if a decision-maker is wondering whether there’s a correlation between two factors in the data, they wouldn’t need to wait for a data scientist to run a regression analysis for them—they could just type their question into Watson Analytics and see the correlation instantly. And it could even spark new ideas, because Watson Analytics would automatically show them other correlations that they might not even have considered. That’s the kind of cognitive approach to analytics that could really help us and our clients out-think competitors.”
Driving swift reactions
By enabling more dynamic and tailored campaigns, IMM expects that the IBM solutions will not only help it to deliver significant savings to client but also increase in value over time.
Sean Baker says: “With IBM Analytics, we can find out whether a particular approach is working by the next morning, rather than waiting until late in the day. This can translate into tens of thousands of dollars of savings, which we can pass on to clients to reinvest for more sales. By continually optimizing our campaigns with IBM solutions, we can drive incremental sales and stretch our clients’ advertising dollars further than ever before—the campaigns we run become more valuable each day than they were the day before because of these applied learnings.”
Today, IMM has dramatically reduced the time its analysts spend compiling reports. Ken O’Brien comments: “Reports that used to take more than 12 person-hours per day to produce are now ready in just 45 minutes. Better still, the process is almost completely automated, which gives our analysts more time to do what they do best—analyze!
“The IBM solutions are helping our analysts be more productive and creative in their work—which translates to greater job satisfaction, as well as helping us deliver a better service and results to our clients.”
With the ability to track and quantify the impact of its digital marketing campaigns, IMM can fulfill its brand promise of delivering measurable value to clients—and the prospect of leveraging IBM cognitive solutions in the future sets a roadmap for even greater innovation and value in the long-term.
Adam Edelman, Founder and CEO, concludes: “The drive for constant improvement is in IMM’s DNA—every campaign, tactic or strategy needs to offer value that increases over time. Our partnership with IBM plays a central role in enabling us to meet this objective, turning our analytics department into a profit center for the business that offers revolutionary results for clients.”