Demand generation systems—products to attract and nurture leads before they are turned over to the sales department—are probably the fastest growing type of marketing software.  This means there is a good chance you will be asked to help select such a system in the near future.  If that happens, what’s the best way to proceed?

First, recognize that many demand generation products are offered on a “software as a service” basis.  This means you won’t be installing, managing, upgrading or customizing the software, so you can skip many of the usual technical considerations.  Of course, you’ll still need to assess security, reliability and scalability.  You may even add some new technical issues, such as integrating with the system as a Web service or via its Application Program Interface.  And there are always business issues including vendor stability, pricing, contract terms, implementation and support.

But mostly you will focus on user requirements.  This can be exceptionally difficult for demand generation systems because they are new to an organization.  Without previous experience, users find it hard to define their needs.

One thing you don’t want to do is just pick the system with the most features.  Features you don’t need add complexity without increasing value.

A better approach is to steal a page from the developers’ book and define business use cases.  Most demand generation systems are used for a few common applications.  It’s easy enough to list these and have your users pick the ones that are most important.  The requirements for these applications are the requirements for your system.

Here is a set of demand generation applications for your marketers to consider.

- Lead generation campaigns.  These campaigns attract new names to your prospect database and reactivate old ones.  They send messages through list-based promotions such as email, direct mail and telephone calls, and anonymous media such as print and Web advertising, search engine listings, and trade shows.  Responses are captured and reacted to appropriately.

Functional requirements for lead generation include campaign setup with costs and tracking codes; list segmentation and selection; creation and delivery of email, direct mail, and telemarketing scripts; landing pages and surveys to capture responses; auto-reply mechanisms to react to inquiries; and reporting on results, including sales data from CRM or accounting systems.

Specific requirements depend on your business.  If your marketers execute many programs simultaneously or produce localized versions, they’ll need features to share standardized content.  If they offer many different products or gather detailed prospect information, look for ability to tailor messages by list segment.  If they work in channels beyond email and Web pages, the system must support these.  If they import leads and sales data from external sources, they need customer data integration and extensible data models.  Demand generation products vary greatly in all of these areas.

- Lead nurturing.  This is maintaining continuous contact with leads until they are ready to buy.  It involves sending messages to inform leads about the company, and gathering information about the leads so you can understand them better.  Functional requirements include multi-step, cross-channel campaign flows; visitor identification through cookies, URL strings and IP look-up; offer selection based on lead attributes and observed behavior; generation of email and newsletters; response capture via surveys and landing pages; and results measurement.  Because lead nurturing programs seek to enrich existing relationships, capturing detailed data is especially important.  So is the ability to use this data in complex program flows, offer selection logic and message personalization rules.  Many lead nurturing programs also need a continuous supply of content to keep the messages fresh, which implies requirements to manage content libraries and RSS feeds.  Companies vary greatly in how sophisticated they need their lead nurturing programs to be.  In general, companies with many products or complex sales cycles will want more advanced lead nurturing capabilities.

- Lead scoring and distribution.  This manages the actual transfer of leads to sales.  Key functions include gathering data through surveys, data enhancement and activity tracking; lead scoring calculations to identify when the leads are ready to send; lead assignment rules to determine which salesperson should receive the lead; and CRM system integration to make the actual transfer, synchronize marketing and CRM data, and coordinate future contacts.  Demand generation systems vary widely in all these areas.  In particular, don’t assume a system can meet your needs if you must: integrate with a CRM system other than Salesforce.com; assign leads within the demand generation system rather than relying on the CRM system; use a Web services interface to interrogate external sources such as D&B or Jigsaw; or perform sophisticated lead scoring calculations.  Also look closely if your salespeople expect true real-time integration, such as alerts when a prospect visits the Web site or requests a chat.

The list of common demand generation applications also includes marketing performance measurement, managing events such as Webinars, and coordination of local marketing efforts.  Each has specific requirements that you can identify by working with your marketers to define the necessary process and data flows.  Once you’ve done this, it’s easy enough to look for the required capabilities in the systems you evaluate.  The trick is making sure the requirements are based on specific business needs: it’s the only way to get the system that’s right for you.

*                            *                           *

David M. Raab is a Principal at Raab Associates Inc., a consultancy specializing in marketing technology and analytics.  He can be reached at draab@raabassociates.com.

Last month’s column described the importance of lead scoring within “demand generation” systems. But perhaps we should step back to describe those systems in general. Many people still confuse them with “marketing automation” or “campaign management” products.

It’s an easy mistake. Both sets of systems maintain a contact database used to drive outbound marketing campaigns. Both provide reporting and analysis tools to understand promotion results. Both sometimes include marketing planning, content management, and project management.

The obvious difference is that demand generation software is nearly always used by business marketers, while marketing automation and campaign management are used primarily to reach consumers. But this distinction is less useful than it seems. Many traditional marketing automation systems are also used for business marketing. And many small consumer marketers use lower-end demand generation software.

The more meaningful distinction is probably between companies that market directly to their customers and those that sell through sales people. Traditional marketing automation systems are used primarily in financial services, travel, retail and communications companies. Their campaigns sell specific products, even though the sale may be completed at a retail store, bank branch or sales agent. By contrast, demand generation systems attract and nurture leads which will be handed to sales departments when they are ready to buy. The salespeople will then identify needs, select appropriate offers, and close the deal.

Another, even simpler difference is that demand generation systems work with leads—that is, people who have not yet made their first purchase—while marketing automation systems focus on existing customers.

The fundamental distinction between nurturing leads and managing customers drives the major differences between the two sets of products. These include:

- focus on Internet behavior. Demand generation systems drive prospects to the company Web site, monitor their behavior, and infer when they are ready to buy. Most of herding is done with emails, which themselves can report whether they have been received, opened, clicked on, etc. Demand generation systems track Internet behavior in great detail because it’s one of their two main information sources. (The other is user-provided information such as surveys). By contrast, marketing automation systems work primarily with promotion and purchase histories. Non-purchase behaviors such as Web site visits are given much less weight if they are considered at all.

- integrated Web pages and analytics. Demand generation systems provide tools to build Web surveys and microsites and to capture data from these directly. This reflects their focus on online media. Marketing automation systems can sometimes build Web pages, but they largely assume this will be done externally. Similarly, they usually rely on third-party Web analytics systems to capture information about visitor behaviors.

- tracking of anonymous visitors. Tagging anonymous Web site visitors with cookies, building a history of their behavior, and later merging that history with the visitors’ identities are central features of demand generation systems. Marketing automation systems may not even track anonymous visitors, and certainly do not consider this a core capability. Have I mentioned that they are primarily interested in communicating with known customers?

- multi-step, highly reactive campaigns. Treatments within a demand generation campaign can vary quickly and significantly in response to an individual’s behavior. Marketing automation systems consider this an advanced feature that only their most sophisticated users are expected to deploy. In contrast, this is a fundamental capability for even basic demand generation products. In fact, finding ways to simplify deployment of multi-step campaigns is one of the main competitive battlegrounds in the industry.

- limited segmentation. This is the flip side of campaign complexity. Demand generation systems start with limited information about their targets, so they build campaigns that adjust treatments as information is gathered during execution. Marketing automation systems begin with a much richer customer history, so they select treatments using complex segmentations when the campaign is set up.

- lead scoring. Demand generation systems support elaborate scoring calculations to measure when a lead is ready for sales. Although marketing automation systems often support user-defined calculations and predictive modeling, they lack specialized lead-scoring functions such as depreciating the points allocated to older events or capping the points generated by a particular type of event. This is another competitive arena for demand generation vendors.

- simple database structure. Both demand generation and marketing automation systems maintain databases with information about individuals. But the base structure in demand generation systems is usually just a lead table and contact history. A modern marketing automation system nearly always includes purchases, and often additional information such as account balances and customer service interactions. The theory among demand generation vendors is that detailed information will be kept in the company’s customer management systems. However, most demand generation systems do let their clients extend the data structure through custom tables. For example, some version of purchase history is needed to measure campaign return on investment.

Many of these differences are more a matter of emphasis than fundamental technology. The products within each group also vary widely. So you still need to identify your own business requirements and assess how each system would meet them. But understanding the distinction between the two categories should make it easier to narrow in on the products best suited to your needs.

* * *

David M. Raab is a Principal at Raab Associates Inc., a consultancy specializing in marketing technology and analytics. He can be reached at draab@raabassociates.com.

There is no hotter category in marketing software today than demand generation systems. These products, from vendors such as Eloqua, Vtrenz, Market2Lead and Marketo, generate and nurture leads through a combination of email, Web pages, lead scoring, and integration with sales automation systems. As with any software, it usually matters less which product you choose than whether you use it effectively. But there are certainly differences among the available products, and picking a system that fits your needs is an important step on the road to success.

So, how do you make a sound choice? One thing not to do is to look at somebody else’s list of the “top three” products or “industry leaders”, and refuse to consider anything else. Industry rankings are important to insiders, but they say little about any product’s relationship to your own business situation. If you were buying a car, would you only look at the three best-selling autos?

On the other hand, few marketers have the skills or inclination to perform an in-depth technical analysis of the several dozen demand generation systems on the market today. Nor do they need to. It’s more important that they document their own needs, but even these are often defined broadly because there’s no time for anything more. What’s needed is a shortcut that lets marketers quickly identify a set of vendors that are all fundamentally suitable. Marketers can then choose how much additional effort to put into selecting the best within this group, trading the gains from an incrementally better fit against the costs of gathering more information.

Such a shortcut is available. Basically it requires identifying the few key differences among systems that relate to your critical business requirements. These fall into three major categories.

1. Marketing Scale. Enterprises where dozens of marketers run hundreds of programs have unique needs. Marketing activities in these large enterprises are often managed separately for different regions, products and customer segments, but must still be coordinated to maintain consistent messages and performance metrics. If you work in a large marketing organization, critical features to look for include:

- modular marketing materials, such as templates with standard headers and footers, reusable Web forms, and automated selection of message text within a promotion. Such features allow a single change to be deployed instantly across multiple campaigns, saving time and ensuring consistency. Without them, a large marketing organization quickly descends into chaos.

- fine-grained security that allows different users to control different marketing campaigns or components. This is not a concern in small departments where one or two users do everything, but it becomes critical where responsibilities for execution are divided.

- ability to run on company systems. Large companies are especially prone to insist that the software run on their own servers, rather than the usual approach (for demand generation software) of letting the vendor run it for them. Or, a company may accept external hosting but want Web forms to be embedded within company-hosted Web pages rather than being part of vendor-hosted “microsites”. Unlike the previous two items, this is only important for some large companies—but for them, it can be an absolute requirement.

- support for languages other than English is something you either do or don’t need. Not all systems have it. ‘Nuff said.

2. Channel Scope. All demand generation systems support emails, Web forms, and integration with CRM systems such as Salesforce.com. If those are the only channels you use, any system will do. But if you work in other channels, especially within the same campaign, you’ll want to find a system that can handle those as well.

- direct mail, telephone, mobile (SMS) and fax are easy if all you want to do is send non-real-time messages in periodic batches. Any system can generate the necessary lists. But if you want an immediate response or something highly personalized you’ll need to look more closely at what each vendor can do and what delivery partnerships it has in place.

- events such as Webinars and seminars require specialized features such as waiting lists, reminders, and attendance reports. Only a few demand generation vendors do this well. But there are plenty of specialist vendors for event management, so this may not be a make-or-break requirement.

- RSS feeds are an increasingly popular way to update content automatically. Simply displaying the feeds is no problem, but some demand generation products can actually manage feed creation and track readership in greater detail than standard RSS technology. Few marketers are so dependent on RSS that they would consider this a core need, but it’s definitely something to consider.

3. Functional Scope. The full set marketing automation functions includes planning, project management, content management, campaign management and analysis. Just two of these (content and campaign management) are essential for demand generation. If you have good solutions in place for the rest, stop right here. Otherwise, decide whether you want a demand generation system that can do:

- planning, which includes budgets and campaign schedules. Integration with corporate accounting systems is desirable but hard to find.

- project management, which includes tracking the tasks to complete a campaign and managing workflow such as approvals. These features are also barely present in today’s demand generation systems, although they will become more common.

- analysis, which includes reporting on Web activity, email delivery, campaign response, return on investment, and much else. Demand generation systems vary greatly in this area, so look closely if you have significant unmet needs. In particular, recognize that comprehensive reporting may require the demand generation system to import and merge data from other sources, which some products simply cannot do.

This brief list is not intended as a comprehensive catalog of possible demand generation system features. Nor does it even touch on other important considerations such as ease of use and vendor stability. What it does offer is a quick checklist of items that are easy to judge against two criteria: (a) do you need it? and (b) does a system have it? The answers will tell you which products are worth exploring in depth.

* * *

David M. Raab is a Principal at Raab Associates Inc. www.raabassociatesinc.com, a consultancy specializing in marketing technology and analysis. He can be reached at draab@raabassociates.com. He is currently researching a Guide to Demand Generation Systems scheduled for publication by September 2008. Mr. Raab also blogs at Customer Experience Matrix http://customerexperiencematrix.blogspot.com/ and MPM Toolkit http://mpmtoolkit.blogspot.com/.

It’s long been clear that relational databases are ill-suited for analytical processing. The problem is fundamental: relational databases are designed to retrieve all elements from a few records, while analytical queries typically read all records but only a few elements. Until recently, database developers were able to overcome the problem with clever design and powerful hardware. Today, the largest databases—hundreds or thousands of terabytes—are too big for this to work at an acceptable price. So the quest for alternatives is on.

But let’s back up a bit. Relational databases are inefficient at analyzing all size databases, not just the largest. Companies may have met their critical needs by throwing money at the problem, but they have also rejected less important applications which couldn’t justify the cost. This suggests there is a significant hidden demand for analytical applications of all sizes if costs can be reduced.

The insight is important because many analytical technologies do not handle very large databases. They are sometimes dismissed for this reason. But this is a mistake: if alternative technologies are more cost-effective than conventional databases for smaller scale projects, they still add value in those situations.

With that in mind, let’s look at the major technologies available today for analytical systems.

- multidimensional databases. These systems read cubes of aggregated data. This makes them fundamentally different from the products listed in the rest of this article, which give direct access to the raw details. This is a critical distinction for analytical applications where the required queries are not known in advance. So now that we’ve mentioned products like Business Objects (www.businessobjects.com), Hyperion (http://www.oracle.com/hyperion), Cognos 8 (www.cognos.com) and Applix TM1 (http://www.cognos.com/applix/) – and acknowledged that they have long been the primary alternatives to relational databases for analytical purposes – let’s move ahead..

- in-memory databases. These include Panoratio (www.panoratio.com), PivotLink (www.pivotlink.com), and QlikView (www.qlikview.com). The technical details differ, but in general all these systems load the source data into memory in a compressed, non-relational format and query against it. Part of their value comes from the compression itself, which lets them handle more data simply loading relational tables into memory. But compression varies greatly with both the technology and the data and is sometimes little better than 1:1. At least much value comes from the non-relational structures, which can be more flexible and powerful at analytical queries than SQL-based systems. This makes the alternative systems cheaper to deploy because there is less fine-tuning of data structures. Still, conventional servers rarely run more than 32 or 64 gigabytes of memory. This means that, even with compression, a pure in-memory system would almost never hold more than fifty or one hundred gigabytes of source data.

- simple columnar systems. These are “simple” in the sense of running on conventional servers, not the massively parallel kind. Vendors include database marketing stalwarts Alterian (www.alterian.com) and smartFocus (www.smartfocus.com). These products organize data by columns (i.e., data elements or fields) rather than rows. Because analytical queries typically read a few elements in all records, the columnar structure lets the database engine load only the information it needs. This reduces the amount of disk access and speeds the result. Like the in-memory systems, these tools have query languages that are more flexible than SQL, making them easier to work with. Scalability is still limited, however: these systems rarely exceed a terabyte of source data.

- massive columnar systems. This is the category receiving the most attention today. Vendors include Calpont (www.calpont.com), EXASOL (www.exasol.com), ParAccel (www.paraccel.com) and Vertica (www.vertica.com). Each is unique but these systems share several basic characteristics: SQL compatibility, massively parallel “shared nothing” hardware, and some flavor of a columnar data structure. These vendors generally supplement the columnar approach with other techniques, such as partitioning and storing multiple copies of the same information in different sort sequences. Some support in-memory data storage as well. These systems do scale, although it seems that most implementations are less than 10 terabytes.

- database appliances. These are also massively parallel systems with a SQL-compatible database. Rather than a columnar approach, they focus on innovations to get the best performance from their hardware at volumes into the hundreds of terabytes. Competitors include DATAllegro (www.datallegro.com), Greenplum (www.greenplum.com), Kognitio (www.kognitio.com) and Netezza (www.netezza.com). The claim here is scalability at a much lower cost than MPP leader Teradata (www.teradata.com).

- extraction-based systems. For tasks like monitoring Web traffic, data volumes are so huge that it makes more sense to scan traffic and extract selected attributes than to store it all. Products including Altosoft (www.altosoft.com), Skytide (www.skytide.com) and Visual IQ (www.visualiq.com) offer this capability.

- outliers. There are a few analytical systems that don’t fit into any of the above categories. illuminate Systems (www.i-lluminate.com), InfoBright (www.infobright.com) and QD Technology (www.qdtechnology.com) work in the sub-terabyte range. All are SQL-compatible, disk-based rather than in-memory, and run on standard servers. illuminate uses a very flexible database approach of its own devising: essentially, it stores each value once and attaches index structures that show all the contexts in which that value appears. Infobright and QD Technology employ a conventional relational structure and focus on compression. Sybase IQ (www.sybase.com) is a variation on a columnar database that reaches very high volumes (it cites a test with 155 terabytes of input) without massively parallel hardware.

* * *

David M. Raab is a Principal at Raab Associates Inc., a consultancy specializing in marketing technology and analytics. He can be reached at draab@raabassociates.com.

Rethinking the Role of CRM Systems

by David M. Raab

Curtis Marketwise FIRST

June, 2008

Hoping to teach their over-optimistic child about life’s grim realities, his parents lock him overnight in a room filled with manure. But the next morning, they find him digging enthusiastically through the muck,  happy as ever. “What are you doing?” the parents ask in exasperation. “With all this mess,” the bright-eyed child calls back, “there has to be a pony in here somewhere.”

Make a few changes, and the same story applies to the masses of data that bankers now collect about their customers. Not so long ago, even the most basic customer information was hard to come by: there wasn’t much to begin with—often just account balances from line of business systems—and even that was difficult for marketers to access. But today, data from line of business systems, background information from external sources, and behavioral details from Web site and call center visits are all easily assembled in a data warehouse or marketing database. The trick is no longer building the pile of data, but sifting through it to find the banking equivalent of that pony.

One version of this sifting process involves sophisticated analytical software to determine which data patterns to look for, and then to find those patterns as they occur. This is a challenging task, since there are many patterns to consider, the details vary from bank to bank, and the patterns themselves can shift over time. But products including SAS Interaction Managment, Unica Affinium Detect , Fair Isaac OfferPoint, Harte-Hanks Allink Agent, Eventricity and ASA Customer Opportunity Advisor all provide the necessary capabilities. Most draw on the vendors’ experience with previous clients to provide a starter set of useful patterns and rules as well. While the technical details vary, each system creates lists of customers whose behaviors match patterns that have been identified as significant. Those lists are handed over to a Customer Relationship Management (CRM) system for action.

Another type of sifting happens when the CRM system itself monitors customer activities for specified conditions. This usually looks for simple conditions, such as visits to the mortgage calculator on a Web site, rather than multiple events over time. The result is still a list of customers to contact, either via phone call, email or direct mail.

Yet another version of this sifting happens during actual interactions, when the CRM system uses its record of previous behaviors to help select customer treatments. This may involve nothing more than displaying a summarized list of past behaviors to a banker talking to the customer in person or by telephone. Or, more powerfully, it may recommend specific treatments based on an analysis of those behaviors and other customer data. For automated systems such as Web sites, ATM machines, or telephone voice response units, it may select the actual treatments themselves.

What all these approaches have in common is that the final customer contact is managed in the CRM system. This means that the CRM system, not the behavior detection technique, is the critical link between assembling masses of customer data and getting business value from that data. Banks can and do apply multiple techniques to identifying opportunities within this data, but it’s up to the CRM system to combine these inputs and ultimately determine what the customer actually sees. The CRM system can now be thought of as a “treatment delivery system”.

This is a relatively new role for CRM systems. Their original function was to themselves act as central repositories for customer information and account history. They presented this information to bankers in call centers and branches so they could answer customer questions and make decisions based on relatively complete knowledge. If they provided any guidance regarding specific treatments, it was usually based on campaigns assigned to customer segments or business rules embedded in telemarketing scripts.

This new role imposes new requirements on the CRM software. It must be more open to inputs from external sources, whether accepting leads identified by the behavior detection systems, capturing non-transactional behaviors from a Web site, or accepting recommendations from a predictive modeling system. It must arbitrate among recommendations provided from the different systems, ensuring that customers are treated consistently over time and across channels, and that the most valuable treatments are chosen among the options presented. It must support an ever-growing array of delivery media, seamlessly merging new channels like mobile Web, video and text messages with traditional call center, branch automation and direct mail. Finally, it must report back to the various recommendation systems, telling them what treatments the customer was actual given and how the customer responded. This feedback is crucial for helping the recommendation systems to improve their own performance.

Older CRM systems may not meet these conditions. Many were designed with rigid data models tailored to the specific needs of call centers and sales automation. Even adding support for Web sites and email can be difficult, while real-time integration with recommendation systems can be nearly impossible. Often the systems were extensively customized during their initial deployment to fit the bank’s technical environment and business processes, making additional changes costly and difficult.

Newer CRM systems are generally more flexible, so this is one area where late adopters have an advantage. One word of caution: the latest rage in CRM software, “on demand’ systems that are run by a third party, should be evaluated very carefully in this area. Although their developers have worked hard to make them more flexible, they may still be too limited for a multi-channel, bank-wide deployment.

In some cases, it may be possible to supplement rather than replace an existing CRM system. For example, vendors including eglue and Infor read data from the CRM system and other sources as an interaction takes place, and then deliver recommendations that can be viewed in a separate window. This allows them to control treatments across multiple channels with minimal changes to the channel systems themselves.

Realistically, many banks today will not be able to afford a comprehensive treatment delivery system. Even so, they should still be able to deploy simpler technologies that monitor some types of customer behavior to identify significant business opportunities. Feeding these as leads into an existing CRM or sales automation system can provide substantial value with minimal technical effort. You already have those huge piles of data—so you might as well start digging.

* * *

David M. Raab is a Principal at Raab Associates Inc., a consultancy specializing in marketing technology and analytics. He can be reached at draab@raabassociates.com.

]]> http://archive.raabassociatesinc.com/2008/06/347/feed/ http://archive.raabassociatesinc.com/2008/06/technical-measures-for-data-quality-investments/ http://archive.raabassociatesinc.com/2008/06/technical-measures-for-data-quality-investments/#comments Sun, 01 Jun 2008 11:56:32 +0000 http://archive.raabassociatesinc.com/?p=54 Technical Measures for Data Quality Investments
David M. Raab
DM Review
June 2008