Apr 08, 2011 (08:04 PM EDT)
IT And Marketing: How Digital Media's Changing The Relationship
Read the Original Article at InformationWeek
Modern marketers don't make emotional decisions in the kinds of boozy, smoke-filled rooms seen on Mad Men. They want some statistical assurances before they spend one red cent, for which they're relying on technology tools and expertise.
Here's the latest formula: Add social, mobile, Web, and email to conventional marketing channels, including print and broadcast advertising, billboards, telemarketing, and direct-mail campaigns. Target and test desired segments, like most-profitable customers. Measure results across all channels. Spend accordingly.
That's a gross simplification, and the basic approach isn't new. Marketers have been targeting, testing, and measuring direct mail and telemarketing for decades, and online advertising and email campaigns for a dozen years.
But there are new channels. Social networks, for example, have introduced direct customer interaction, and they're making it easier to measure broadcast campaigns. Agencies and advertisers, for example, now routinely monitor YouTube and Hulu activity to see which ads are most popular and which moments within those ads are being replayed. They also monitor Facebook and Twitter to get a sense of positive and negative sentiments. And the latest smartphones add the dimension of location awareness to Web and email advertising.
The thing that has changed most in recent years is that marketers want a more comprehensive, cross-channel examination of campaign results, and they expect more detailed analyses. It's now all about the data, so there are good employment prospects for IT pros who know how to manage and analyze that data. Some of the biggest IT vendors see this opportunity, and are snapping up marketing software specialists.
The bad news is that marketers are still frustrated by gaps in measurability across channels. And while they're trying to fill those gaps by buying software to manage and analyze advertising efforts, marketing execs don't necessarily trust their own IT departments to implement it.
It's high time for IT pros to be working more closely with their marketing colleagues to understand marketing's needs and goals. If IT can't deliver the requisite data management and analytic modeling, the work will end up moving to the marketing team or to a third-party service provider.
Statistical rigor has been applied to marketing for years, but there are still far too many siloed applications--for tracking and executing direct-mail campaigns, email campaigns, Web analysis, lead nurturing, print and broadcast ad buying, and more.
Marketers have yearned for a holistic, cross-channel view of campaigns and spending. A crop of vendors responded in the 1990s with software known variously as marketing automation, marketing resource management, enterprise marketing management, and integrated marketing management. These software products have recently come to the attention of mainstream IT vendors. Oracle acquired the technology behind Market2Lead in May 2010, IBM acquired Unica in October 2010, Teradata bought Aprimo in December 2010, and SAS acquired Assetlink in February 2011.
These acquired companies have different strengths: lead nurturing in the case of Market2Lead, e-marketing at Unica, planning and financial management at Aprimo, and workflow and digital asset management for Assetlink. What they all have in common is marketing-specific capabilities and a focus on cross-channel analysis.
IBM, Oracle, Teradata, and SAS are diving more deeply into marketing because "they see it as a huge analytical play," says David Raab of Raab Associates, an analyst who specializes in marketing technology. "We've been talking for years about coordinating all the customer contacts, but now big IT vendors are starting to put some muscle behind the concept." Marketing is looking like one of the next great opportunities in analytics, following on successes in areas such as fraud detection, supply chain optimization, and risk assessment.
Justify The Budget
A year after starting an enterprise-wide initiative to let the data drive its marketing decisions, Kodak is starting to see the effects. The company's approach used to be top down: set up annual budgets for broadcast, print, direct mail, online, email, and other marketing programs based on last year's spending. It now tries to constantly forecast the lift expected from each marketing spending, with measurement guidelines for each channel. "We're still going to give you a budget, but you have to show us directly how you're going to influence revenue," says Dale Brosch, Kodak's manager of CRM for the Americas.
Measurement is a given with direct mail, online, and email campaigns, and Kodak does extensive A/B comparisons that isolate the effect of ad creatives, promotional offers, customer segmentations, and messaging across all channels. It measures print ads and retail circular promotions, too, albeit more slowly by matching product registrations, which ask when and where customers purchased, against records of campaigns. About three in 10 customers register their products, so Kodak extrapolates as scientifically as it can.
As for the newest channels, Kodak employs a chief listener and a chief blogger, both of whom monitor and respond to relevant social network comments and customer reviews. Their outbound efforts must dovetail with the messaging and communications strategies developed for all other channels. And their success in influencing social channels is measured--with help from social network review platform BazaarVoice--and reported up to the chief marketing officer.
For customer segments that use smartphones heavily, Kodak tracks what kinds of devices people use to access an email or website, and it customizes campaigns for the most popular platforms, since the experience can vary device to device. "If you don't design for mobile, you don't know what customers are seeing on the device," Brosch says.
Kodak's customer segmentation is much more fine-grained than it used to be. The manufacturer formerly built campaigns around customer personas that were "more aspirational than real," Brosch says, citing the proverbial 25- to 34-year-old suburban soccer mom with two kids.
Kodak now leans more on its database, which has information on some 50 million consumers, aggregated from product registrations, customer-service interactions, direct purchases on the Kodak retail website, purchases and registrations on its photo-sharing site, and interactions through social networks. Kodak appends psychographic and demographic information to this data from third-party databases, a common practice for marketers.
All this data has allowed Kodak to replace those crude personas with in-depth profiles of customer groups and buying patterns. For example, buyers of the entry-level C310 multifunction printer have a very different profile from those who buy the more expensive ESP Office 6150. "We skew everything, including the retail packaging, the product descriptions, and the ad creative content for every channel, to what we know about those consumers," Brosch says.
Fine-grained segmentation is made possible by analytics, and it drives more cost-effective purchasing of lists and media placements. Instead of advertising the entire ink-jet printer line in a broad publication, for example, Kodak will target probable 6150 buyers though a lower-cost combination of print ads in niche magazines, Web banner ads on selected sites, and email lists known to reach small-office or home-office buyers.
The basic outline is old hat, but Kodak and other marketers have had to expand and improve their data aggregation and management capabilities to unite formerly separate silos of customer information. They've also had to refine their customer segmentation and then test across more channels.
Manage The Marketing
Making the most of customer information is more than half the battle, but marketers are also starting to consolidate all the data on their marketing activities. That's where the aforementioned marketing systems enter the picture. In place of a hodgepodge of spreadsheets, these systems track campaign plans, budgets, customer targets, lists used, and results, and in some cases even track performance of individual content components such as ad copy, images, video clips, and variations by language.
Kodak uses software from Aprimo, the company acquired by Teradata, to manage campaigns and customer segmentation on the business-to-business side of the house, which includes its commercial printing, document imaging, and TV and motion picture divisions. The software taps into Kodak's customer database to pick B-to-B customer segments to target with ads via Web, email, and mobile devices. As for Kodak's business-to-consumer operations, it's evaluating whether to go with Aprimo in-house or outsource to a data management company like Epsilon or Acxiom, which have sophisticated tools for segmentation and data mining. They also have integrated email and mobile support, so once the analysis finds the right customers to target, companies can turn that right into email campaigns and advertising tailored to the iPhone, BlackBerry, and most Android devices.
The Big Vendors Are In
Teradata acquired Aprimo to extend its reach into marketing departments. It already had integration partnerships with Aprimo, whereby the marketing tools could tap Teradata data warehouses and marts. Kodak's customer database is currently Oracle, and no doubt Teradata hopes Aprimo's relationship with Kodak and other customers will open doors to future data warehouse deployments.
If Kodak turns to a marketing service provider like Epsilon or Acxiom for more of its data analysis, it likely would need fewer staffers to do such work. It also would introduce questions about how to integrate data between the on-premises data sources and those held by the service providers, and responsibilities for maintaining and securing customer data.
It's similar to the kinds of IT resource decisions most companies will face in the years to come as cloud computing takes hold. Which data and capabilities do you really want on premises rather than in the cloud? What infrastructure is too much bother or expense to develop and maintain internally? What level of sophistication can you hope to achieve internally versus what's available as a service?
Unica, Aprimo, and Assetlink support cloud-based as well as on-premises deployments, and there are plenty of signs that IBM, SAS, and Teradata are preparing for a hybrid world. IT pros need to do the same, but more on that later.
Measure Online Impacts
Royal Bank of Canada (RBC) is capturing and managing customer data in-house, but in 2005 it added an on- demand marketing operations management service from Assetlink, the vendor recently acquired by SAS, to track its marketing efforts.
"We didn't have good operational metrics about what campaigns were done, when they were done, their cost, which strategy they supported, and whether they were integrated across multiple channels," says Lorraine Hand, director of marketing operations at RBC.
Consolidating data that was previously scattered across spreadsheets and executive PCs, the RBC Assetlink system now has more than five years of information on strategic marketing plans, budgets, and campaigns across broadcast, cable TV, direct mail, email, and telemarketing. In the past year, RBC added detailed data on ad placements, changes in brand metrics (tracked by surveys), and campaign responses.
The Assetlink database has grown so large that Hand says it likely will be integrated with RBC's data warehouse. Companies are finding that targeting and campaign result data has to be accessible alongside customer information, but that doesn't mean RBC has to move Assetlink to an on-premises deployment. The Assetlink data can be treated as just another data source--albeit a rich one--as far as integration is concerned, so the usual batch or Web service data integration options should do. As long as the data's accessible, marketing can do its analysis.
RBC's goal is to understand what's working and what isn't, and the company goes to great lengths to track variations. For example, RBC may employ as many as 200 toll-free numbers and URLs to test and track a large credit card campaign across cable TV, online, email, direct mail, and even print and airport billboard advertising. RBC can also measure the lift from broadcast ads--what Hand calls "air cover," which the bank used extensively during last year's Winter Olympics in Vancouver.
"It's about looking at the whole marketing mix, optimizing it, and making sure that it's worth putting the dollars against more expensive TV campaigns," Hand says.
Gaps In Analysis
The one area where RBC isn't satisfied is online advertising. That might seem like the channel that's most measurable, but RBC can't connect clicks to completed sales in most cases, in part because of regulatory hurdles that most American companies don't face.
The problem with online is that there's almost too much data, Hand says. Yet because of Canada's tough know-the-customer restrictions related to opening financial accounts online, it's hard to tie clicks to new business. Financial services companies can sell simple products online like credit cards, but only to existing customers. New bank accounts and almost any product sold to a new customer requires a phone call or branch visit, Hand says. So did someone call to start an account because they saw an email ad, visited the website, or even saw a poster?
"We may know that we drove somebody to click on a URL and five weeks later we may see an increase in our portfolio," she says, "but it's hard to connect that click to somebody picking up a phone or walking into a branch to buy a product." RBC is experimenting with reporting and analytics software to determine which online campaigns were running, when they were running, and how they influenced sales of financial products.
Scale Back Brand Advertising
Under pressure from the recession and high unemployment, job listings site Monster.com scaled back its broad-based, brand advertising in favor of a more cost-effective, multichannel approach. After a bit of experimentation, it has settled on combining personalized email, direct mail, social engagement, and prioritized telemarketing.
Monster tracks all those efforts in a campaign management system from Unica, the vendor IBM acquired last year. The on-premises system is supported by the company's applications team, which meets with marketing weekly to review trouble and service tickets, ongoing projects and changes, and the regular release schedule.
If IT wants to build trust with marketing, this kind of close interaction can be critical. In a survey of marketing and IT teams by the CMO Council and Accenture, nearly two-thirds of marketers say they had problems implementing marketing software, and the No. 1 reason cited was the low priority IT puts on marketing.
At Monster, a typical campaign starts with email, but it has evolved from sending generic messages to big groups to sending personalized messages to targeted segments. Its best prospects are HR decision-makers at large companies in growth industries, such as high tech and healthcare.
Monster uses SAS statistical modeling software to identify customers and prospects most likely to buy job listings and other services in a given quarter. These scores are applied to the marketing database under Unica, which also maintains details on when campaigns ran, who received them, who opened them, and who clicked through. Unica's campaign module pulls mailing lists based on a range of criteria, including the opportunity score and past response behavior.
Cream-of-the-crop HR execs might receive a high-touch--and expensive--direct mail promotion, so the stakes are high for getting the analytics right. For example, Monster recently sent top prospects GPS devices to promote its new Power Resume Search product, since the promotional copy described it as a GPS for finding job candidates. "We use analytics against attributes like location, industry, and company size to score the data and home in on a subset of maybe 1,000 customers where we want to spend that kind of money," marketing VP Matthew Resteghini says.
Monster also has its social media team try to initiate interaction with select prospects through LinkedIn and other business-oriented social networks and groups. These prospects, too, are pulled from Unica. If the target list has 1,000 executives, Monster staffers might engage 50 to 100 of them through social media.
Behaviors tracked in Unica also drive prioritized telemarketing follow-up calls from Monster's sales force. For example, any customer who has opened and clicked through on more than one email is likely to get a call.
Resteghini says Monster is getting much better results now that it presents a consistent, targeted message across its core direct-marketing channels of email, direct mail, and telemarketing. But he's not entirely happy, particularly with measuring the impact of broadcast advertising--something RBC seems more confident about.
Monster also wants to do a better job of integrating banner ad campaigns with its email, direct mail, and telemarketing efforts. Web ads are now tracked through separate Web analytics tools, helping Monster determine how much revenue online campaigns are generating. The company's considering using Unica's NetInsights Web analytics product. With better Web analytics integration, Resteghini says, Monster could also use behavioral data--clicks and other actions tied to online ads--to drive direct-marketing efforts.
The Marketing Arms Race
Marketing technologies are essentially following the same path as many other enterprise technologies have followed: Point software tools are becoming modules in more comprehensive suites, with the promise of easier and tighter integration and improved visibility across activities. That's the theme from big Web analytics vendors Adobe and WebTrends, which have announced their intentions to span social, mobile, and Web analytics.
It's noteworthy that IBM, SAS, and Teradata say they will maintain the brands and standalone business status of their Unica, Assetlink, and Aprimo acquisitions. It's an admission that marketers prefer to deal with specialized marketing firms, not IT behemoths.
IT services firms are targeting marketing analytics as well. IBM, for one, recently started an e-commerce practice that leverages its Unica (integrated marketing), Coremetrics (Web analytics), and Sterling Commerce (order management and supply chain) acquisitions, with a focus on mobile and social networks.
In this transition to digital marketing, IT organizations need to step forward. Many a marketing department subscribed to Salesforce.com without involving IT. In the CMO Council/Accenture survey, only one-fourth of marketers said they consult enterprise IT, contact center, and back office groups when selecting and deploying marketing software. Almost half of the marketers surveyed (46%) don't think their CIO understands marketing objectives.
If IT is going to be relevant, it needs to show it can help marketing teams with information management (integrating, cleansing, and enriching data) and the analytics of modeling and testing results. Marketing has been a data-driven discipline for a long time, but the number of data sources keeps multiplying and the richness and volume of data keeps growing. The IT jobs tied to supporting those efforts will stay in house only if IT teams bring a unique understanding of the company's products and strategy. If not, that work will swing to outsourcers with specialized expertise.
As for the marketers, IT must remember that only one thing motivates their technology choices. "Marketers don't wake up in the morning and say, 'I want to get integrated,'" Raab says. "They wake up and say, 'I want to make money.' They don't want to run cross-channel campaigns because they think they're cool. They do it because [the campaigns] give them a lift on their results."
If IT teams can deliver that lift, through better targeting, more efficient campaign execution, and more measurable results, they'll be invited to the digital marketing table.
'Do Not Track' Would Change The Rules
Download a free PDF of