Wednesday, June 25, 2008

More Blathering About Demand Generation Software

When I was researching last week’s piece on Market2Lead, one of the points that vendor stressed was their ability to create a full-scale marketing database with information from external sources to analyze campaign results. My understanding of competitive products was that they had similar capabilities, at least hypothetically, so I choose not to list that as a Market2Lead specialty.

But I recently spoke with on-demand business intelligence vendor LucidEra , who also said they had found that demand generation systems could not integrate such information. They even cited one demand generation vendor that had turned to them for help. (In fact, LucidEra is releasing a new module for lead conversion analysis today to address this very need. I plan to write more about LucidEra next week.)

Yet another source, Aberdeen Group’s recent study on Lead Prioritization and Scoring: The Path to Higher Conversion (free with registration, for a limited time) also showed that linking marketing campaigns to closed deals to be the least commonly available among all the key capabilities required for effective lead management. Just 64% of the best-in-class vendors had this capability, even though 92% had a lead management or demand generation system.

Quite frankly, these results baffle me, because every demand generation vendor I’ve spoken with has the ability to import data from sales automation systems. Perhaps I'm missed some limit on exactly what kind of data they can bring back. I’ll be researching this in more detail in the near future, so I’ll get to the bottom of it fairly soon.

In the meantime, the Aberdeen report provided some other interesting information. Certainly the overriding point was that technology can’t do the job by itself: business processes and organizational structures must be in place to ensure that marketing and sales work together. Of course, this is true about pretty much any technology, but it’s especially important with lead management because it crosses departmental boundaries. Unfortunately, this is also a rather boring, nagging, floss-your-teeth kind of point that isn’t much fun to discuss once you’re made it. So, having tipped our hat to process, let’s talk about technology instead.

I was particularly intrigued at what Aberdeen found about the relative deployment rates for different capabilities. The study suggests—at least to me; Aberdeen doesn’t quite put it this way—that companies tend to start by deploying a basic lead management platform, followed by lead nurturing programs, and then adding lead prioritization and scoring. These could all be done by the same system, so it’s less a matter of swapping software as you move through the stages than of making fuller use of the system.

If you accept this progression, then prioritization and scoring is at the leading edge of lead management sophistication. Indeed, it is the least common of the key technologies that Aberdeen lists, in place at just 77% of the best-in-class companies. (Although the 64% figure for linking campaigns to closed deals is lower, Aberdeen lists that under performance measurement, not technology.) Within lead scoring itself, Aberdeen reports that customer-provided information such as answers to survey questions are used more widely than inferred information such as Web site behavior. Aberdeen suggests that companies will add inferred data, and in general make their scoring models increasingly complex, as they grow in sophistication.

This view of inferred data in particular and scoring models in general as leading edge functions is important. Many of the demand management vendors I’ve spoken with are putting particular stress on these areas, both in terms of promoting their existing capabilities and of adding to them through enhancements. In doing this, they are probably responding to the demands of their most advanced customers—a natural enough reaction, and one that is laudably customer-driven. But there could also be a squeaky wheel problem here: vendors may be reacting to a vocal minority of existing customers, rather than a silent majority of prospects and less-advanced clients who have other needs. Weaknesses in campaign results reporting, external data integration and other analytics are one area of possible concern. General ease of use and customization could be another.

In a market that is still in its very early stages, the great majority of potential buyers are still quite unsophisticated. It would be a big mistake for vendors to engage in a typical features war, adding capabilities to please a few clients at the cost of adding complexity that makes the system harder for everyone else. Assuming that buyers can accurately assess their true needs—a big if; who isn’t impressed by bells and whistles?—adding too many features would harm the vendors own sales as well.

The Aberdeen report provides some tantalizing bits of data on this issue. It compares what buyers said was important during the technology assessment with what they decided was important after using the technology. But I’m not sure what is being reported: there are five entries in each group (the top five, perhaps?), of which only “customizable solution” appears in both. The other four listed for pre-purchase were: marketing maintained and operated; easy to use interface; integration with CRM; and reminders and event triggers. The other four for post-purchase were: Web analytics; lead scoring flexibility; list segmentation and targeting; and ability to automate complex models.

The question is how you interpret this. Did buyers change their minds about what mattered, or did their focus simply switch once they had a solution in place? I’d guess the latter. From a vendor perspective, of course, you want to emphasize features that will make the sale. Since ease of use ranks in the pre-purchase group, that would seem to favor simplicity. But you want happy customers too, which means providing the features they’ll need. So do you add the features and try to educate buyers about why they’re important? Or do you add them and hide them during the sales process? Or do you just not add them at all?

Would your answer change if I told you, Monty Hall style, that there is little difference between best-in-class companies and everyone else on the pre-sales considerations, but that customization and list segmentation were much less important to less sophisticated customers in the post-sales ranking?

In a way, this is a Hobson’s choice: you can’t not provide the features customers need to do their jobs, and you don’t want to them to start with you and switch to someone else. So the only question is whether you try to hide the complexity or expose it in all its glory. The latter would work for advanced buyers, but, at this stage in the market, those are few in number. So it seems to me that clever interface design—exposing just as many features as the customer needs at the moment--is the way to go.

Wednesday, June 18, 2008

Market2Lead Offers Enterprise-Strength Demand Generation System

Market2Lead offers the usual list of demand generation functions: outbound email, Web forms and landing pages, automated lead nurturing, integration with sales, and campaign return on investment analysis. But while many demand generation vendors simplify these features so marketers can run them for themselves, Market2Lead offers no such compromises. It is designed for world-wide deployment at large enterprises, where even the marketing department will have significant technical resources. This lets the system include capabilities that large installations need and other systems may have skipped.

This is not to say that Market2Lead is especially hard to use. Although the interface is not particularly pretty, the effort to set up a simple campaign is probably about the same as in other demand generation systems. That difference is that Market2Lead is designed to work with other company systems, rather than assuming it will provide almost all of the marketing department’s customer-facing functions. It also provides the administrative muscle needed to manage large marketing programs.

Let’s talk first about working with other company systems, because that strikes me as the most distinctive aspect of Market2Lead. Nearly all demand management systems have built-in email engines and Web servers for forms and microsites. This makes these functions as simple as possible to set up and use. But Market2Lead does not provide its own email services. Instead, it provides an API that lets users set up their emails with Market2Lead and then have them sent by external email specialists like Responsys and Exact Target. Market2Lead’s approach to Web pages is more flexible: users can either build and serve Web templates within Market2Lead, or they can embed Market2Lead tags in externally-hosted pages. These tags call Market2Lead content, including personalized messages and Web forms.

Administrative capabilities to support large marketing programs are the sum of many details. Components including lists, offers and Web forms are stored in libraries where they can be standardized and reused across campaigns. Campaigns and programs within campaigns have start and end dates, which the system can use to automatically adjust the campaign flow after an end date is reached. Users can define standard process flows for outbound campaigns (such as Webinar invitations) and inbound responses (such as inquiries), again to standardize processes and save rework. Offer and campaign priorities are determined by numbers rather than placement in a traditional flow chart, making them easy to rearrange. Suppression lists can be applied automatically to all campaigns, allowing the system to enforce limits on the number of contacts per person. Logic to select questions needed to complete a prospect’s profile is managed centrally rather than embedded separately within each form.

Market2Lead also supports large marketers with multi-language capabilities. It can support programs in 42 languages, store a default language for each prospect, and capture information in different languages from the same person.

Working in yet another dimension, the system extends to channels beyond Web and email. It can manage call center scripts internally, using special versions of its Web forms. Sales force contacts are, of course, captured through two-way integration Salesforce.com and other sales automation systems. Data from still other sources can be merged using data integration tools including a customer matching engine. Users can analyze the resulting marketing database with Informatica reporting tools. There are also standard reports on campaign effectiveness and marketing opportunities. Reports can show how often each list and offer is used and can show the offers and lists used within a campaign. The system show which campaigns use a given list or Web form, but not (yet) which campaigns use a given offer.

Like other demand generation systems, Market2Lead measures campaign effectiveness by importing sales opportunity results from the sales force system. These are tied to marketing campaigns by linking the sales opportunity to prospects, linking prospects to marketing activities, and linking the activities to campaigns. These revenues are matched against campaign costs to calculate return on investment. The system takes a fairly sophisticated approach to this process: it can credit revenue to either the first or last campaign linked to a sale, or share the revenue among multiple campaigns based on user-defined weights. The vendor is researching data mining features to help users determine the appropriate weights.

All of this is impressive, but none of it would matter if the core campaign engine were problematic. But it’s just fine. The basic approach is similar to other campaign managers: users define selection rules for each campaign, and then a series of steps the selected prospects will follow. In the Market2Lead, the selections are expressed as list memberships. These can be frozen at the time of initial selection or dynamically reselected on a regular schedule. Reselection is what lets the system enforces contact limits, since it can continuously recheck whether each prospects has received the maximum number of messages.

Each step within the campaign is defined by an automation rule. These have four components: a “trigger event” such as an email being opened; additional qualifying conditions based on attributes or previous behaviors; a wait period; and a next action to take if the preceding conditions are met.

The first three of these are about what you’d expect. But the next action is more interesting. It can execute the next program within the campaign, exit the campaign, or move the prospect to a different campaign. But the "next" program is not selected from a fixed sequence. Rather, Market2Lead can dynamically select from a list of programs associated with the campaign. Each program has a user-assigned business value and its own qualification conditions. The system can rank the programs by highest, lowest or random value, and will then pick the first program within this ranking that the current prospect is qualified to receive.

You might want to read that last sentence again, because there's a lot going on here. Between the rankings and offer qualifications, a single campaign step can deliver completely different experiences for different prospects without explicit programming. In addition, users can change those experiences without changing the campaign logic itself, simply by altering the business value or qualification rules. Since the same program can be used in different campaigns, a single change can modify many campaigns at once.

One thing the system cannot do is dynamically recalculate the business value itself based on the individual prospect's attributes. Market2Lead is researching this capability.

This approach is admittedly more complicated than setting up a simple campaign in some other demand generation systems. But it all makes sense and will ultimately be more efficient for large marketing operations once you get the hang of it. In any case, Market2Lead says the system is usually run by a marketing operations specialist, who prepares standard program flows and content templates that marketing program managers can then customize for themselves. Training this person and setting up the initial campaign including flows and content templates takes four to six weeks.

As the very notion of a dedicated marketing operations specialist implies, we are talking here about large organizations. Indeed, Market2Lead lists Cisco, Citrix, Interwoven, Netgear, Polycom, SGI, Sungard and Tibco among its 50+ clients. The company does serve smaller organizations as well, and in fact can provide services for clients who do not want to run it for themselves.

Market2Lead offers several configurations of the system with different prices and capabilities. All versions are available as hosted services, and enterprise clients can also install it in-house. Pricing depends on the version, database size and transaction volume and ranges from $60,000 to $100,000 for a mid-market solution. The company was founded in 2003.

Monday, June 16, 2008

Raab on DM Radio Panel on July 10

I'm scheduled to appear on a DM Radio panel on columnar databases on July 10. I'll be joining Dr. Michael Stonebraker, architect of the modern INGRES and POSTGRES database designs, and CTO/Founder of columnar database developer Vertica, and several additional guests.

For more information and to register, visit http://www.dmreview.com/dmradio/10001493-1.html.

Tuesday, June 10, 2008

Marketo Aims to Simplify Demand Generation

As I wrote last week, demand generation vendors have a hard time differentiating their systems from each other. One company that has made a concerted effort is newcomer Marketo. Marketo has a one-word elevator speech: simplicity.

That’s not to say Marketo is a simple product. Functionally, it covers all the demand generation bases: outbound email, landing pages, Web site monitoring, lead scoring, multi-step nurturing programs, prospect database, analytics, Salesforce.com integration. It even adds A/B testing for landing pages, which you don’t see everywhere. The depth in each area is perfectly respectable as well.

Where simplicity comes in is the user interface. Like every other demand generation vendor, Marketo has wrestled with how a branching, multi-step lead nurturing campaign can be made easy enough for non-specialist users. The traditional approach has been a flow chart with lines and boxes. This is, after all, the way “real” process diagrams are built by programmers and engineers. It does express the logic of each flow precisely, but it also can get incomprehensibly complex very quickly.

Marketo’s solution is to do away with the branches. Each campaign flow is presented as a list, and any deviation from the sequence is treated as a diversion to another flow. The list itself can be presented in a collapsed format with each step as a numbered item, or an expanded format where the actions taken at each step are exposed. (Or, users can expand a single step at a time.) Actions include adding or removing the lead from a list, changing a data value or score, sending an email, moving the lead to a different flow, removing it from all flows, and waiting a specified period of time. The system can also add the lead to a Salesforce.com database, assign or change the owner in Salesforce.com, and create a Salesforce.com task. Each action can be associated with a set of conditions that determine whether or not it is executed. One step can include multiple actions, each with its own conditions. The system can be told to execute only the first action whose execution conditions are met, which is one way to implement branching logic .

Other components of Marketo are more conventional, although still designed with simplicity in mind. Users can set up Web landing pages and email templates using a drag-and-drop interface modeled on PowerPoint—the one tool, as Marketo points out, that every marketer is guaranteed to know how to use. These templates can include variables selected from the Marketo database for personalization. Users can also create forms to capture data provided by site visitors or read automatically from the form or URL parameters. Forms can be reused across campaigns.

Campaign lists are built with another drag-and-drop interface, allowing users to layer multiple selection conditions. These can be based on lead data and constraints such as Web search terms, event frequency, and date ranges. Lists can be frozen after selection or dynamically refreshed each time they are used. Users can review the members of a list and click on a name to see its details, including the log of messages sent and activities recorded in Marketo. Like other demand generation systems, Marketo uses cookies to track the behavior of anonymous Web visitors and merge these into the lead record if the visitor later identifies herself. Lead scores are calculated by adding or subtracting points for user-specified behaviors. These values can automatically be reduced as time passes after an event.

Leads can also enter a campaign through triggers. Trigger events can include clicking on a link, filling out a form, changing a data value, creating a new lead record, and being added to a list. The system reacts to triggers as soon as they happen, rather than waiting for lists to be updated.

Campaigns can be scheduled to run once or at regular intervals. So can the wide range of standard reports covering, covering campaign results, email performance, Web activity and lead statistics. Users can run a report against a specified list and can have a report automatically emailed to them on a regular basis. A custom report builder is due by the end of July.

Marketo’s integration with Salesforce.com also bolsters its claim to simplicity. The system feeds data to Salesforce in real time and receives data from Salesforce every five minutes. This will go to real time as soon as Salesforce permits it. The integration is based on the Salesforce Force.com platform, which allows new installations of Marketo to connect with Salesforce in minutes. It also allows Marketo fields to appear within the regular Salesforce tabs, instead of a tab of its own. The lead activity summary from Marketo does appear separately within Salesforce.

It more or less goes without saying that Marketo is sold as a hosted service. This, combined with the automatic Salesforce.com integration, enables new clients to get started very quickly. The company cited implementations in as little as 24 hours, although I’m not sure this is a standard promise. They do say users become proficient after two hours of training. Perhaps the most convincing evidence that the system is easy to install is that the company doesn’t charge a separate set-up fee—definitely not something all its competitors can say.

In fact, Marketo pricing is about as simple as it gets: a straight monthly fee ranging from $1,500 to $10,000 depending on the number of leads, page views and email contacts.

Marketo was founded in late 2005 by veterans of Epiphany. Its leaders spent the first two years researching market requirements and raising capital. They officially launched the Marketo product in March of this year and now have about 35 clients. These are primarily mid-to-large business-to-business marketers.

Friday, June 06, 2008

Oh, the Irony! Do Demand Generation Vendors Have A Sound Marketing Strategy?

There are literally dozens of vendors offering “demand generation” software, which can be roughly defined as systems to generate and nurture leads before turning them over to sales. Their function lists usually sound pretty much alike: outbound email campaigns to generate leads; Web landing pages to capture responses; lead scoring to determine how to treat them; multi-step email campaigns to nurture them; integration with sales automation systems; and analytics to track the results.

It’s a situation that cries out for vendors to specialize in different customer segments, but so far the only division along those lines seems to be that some vendors focus on small businesses while others target mid-size and larger organizations. If anyone has taken the obvious next step of creating vertical packages for specific industries, I haven’t seen it. (Of course, given how many vendors there are and the fact that I do like to sleep occasionally, it’s quite possible that these do exist.)

My guess, however, is that most competitors in this market (and their financial backers) are not yet ready to give up the dream of being the dominant player in a single, unified industry. The generic approach may also reflect the origins of many vendors in technologies like Web analytics and email campaigns, which themselves have not fragmented into vertical specialties. And I suppose the underlying features required for demand generation are in fact largely the same across most segments, so there is little technical reason to specialize.

One implication of this is that resellers, who do tend to specialize by industry, will play an important role in making these systems work for most users. This is what happened with sales automation software. If I’m correct, then the demand generation vendors should be competing aggressively to attract reseller support. I can’t say I’ve seen much of that either—so, again, maybe I’ve missed it. Or maybe deployment of these systems is so simple that resellers can’t make any money doing it. If that were the case—and I’m not convinced it is—then we’d expect vendors who already assist marketers, such as advertising and publicity agencies, to offer the tools as an extension of their own services. I’ve seen a bit of that, but it doesn’t seem to be the main strategy the demand generation vendors are pursuing.

So if demand generation vendors are not staking out vertical specialties or pursuing channel partners, just how do they seem to compete? Many show a reassuring confidence in their own systems, running exactly the sorts of lead generation and nurturing programs they are proposing for others. These are, or should be, coupled with attempts to build superior sales organizations that will close the leads the systems generate. Perhaps this sort of ferocious, nuts-and-bolts approach to sales and marketing is all they really need to win.

And yet, it would be more than a little ironic if companies that hope to sell to marketers were themselves ignoring the strategic marketing issues of branding, differentiation and segmentation. Ironic, but not inconceivable: the demand generation vendors are after all business-to-business marketers, selling primarily to other business marketers, who often pay much less attention to the grand marketing strategies than their consumer marketing counterparts. For better or worse, many business marketers focus primarily on product features and technologies, despite ample historical evidence that the best product does not always win. It should not be surprising that many demand generation marketing programs take a similar approach.

That said, some demand generation systems do seem to have a clearer approach to positioning themselves in the market. This post was going to profile one such vendor. But the preceding introduction ran on for so long that I think I’ll write about that vendor in a separate post next week instead. After all, it’s a Friday afternoon in the summer. Enjoy your weekend.