Marketing Automation and Lead Management Blog

Why is supporting duplicate contacts based on email address important?

Recently, in a conversation with Marketing Automation analyst David Raab, we discussed why anyone would actually want to support multiple prospect records with the same email address in their marketing automation system. This may be a religious war right up there with Captain Kirk vs Captain Picard or Emacs versus VI (ok, maybe I'm a nerd...).

Some marketing automation systems do not support multiple contacts with the same email address.  As a marketing purest I agree with the concept, but unfortunately the buying public does not! Here are three reasons why you will benefit from supporting multiple contacts with the same email address in you marketing automation system.

3 reasons why marketing automation systems should support duplicate email addresses

1. Salesforce.com is notorious for permitting duplicates. Sales people simply add new contacts and leads without hitting the link to check to see if the person already exists in the database. Why is this a problem? Consider this scenario.
   Lead exists in database, synched between Marketing Automation and CRM. Sales gal adds a duplicate contact, same email address.Marketing automation system that only supports one contact per email address cannot synch the duplicate over. Marketing continues to record interactions with the original contact whilst sales records their interactions in the duplicate copy. At some point thecontact tells the rep they want to unsubscribe. Sales gal unsubscribes the non-synching duplicate in CRM.Trouble looms.
2. Many companies think of Leads, not as people, but as early stage opportunities for products. So one person, may represent several leads (interested in several products). In my naive marketing state some years ago I attempted to tell customers what abad idea this was. Telling billion dollar firms to rewire their sales process and CRM, and re-train the global sales team, to accommodate purest marketing concepts...ahhh not such a good idea. This is one of the leading reasons marketing automation systems that cannot support multiple contacts with same email address are often not integrated with a CRM!
3. A slightly different twist on #2 is when different divisions of the same firm, may have contacts who are interested in different products, and have different contact owners. If a person expresses a different product interest by registering for a document or event for a different product, the marketing automation system should be able to create a new contact record for that person, because the CRM will assign ownership to a different individual in a different division. You can still create "fatigue" rules to ensure that the person doesn't get overloaded with email from multiple divisions, but their contact records for each division can be unique. This is considered a very cool feature by some of our Market2Lead enterprise accounts.

Bottom line, life is much better, and data quality is higher if you have a marketing automation system that lets you deduplicate on any combination of fields you choose, but still maintains a separate master opt-out database so that if any of the duplicate contacts opts out at a global level, no email goes to any of the contacts with that email address. It goes without saying that our global customers drove us to develop Market2Lead this way, and we thank them for it! Marketing listening to Sales.

-Kevin

PS sorry I neglected my blogging for a few weeks we were finishing up our new website. Check it out! Lots of lovely new short video demos of the application.

How to stop bad data entering your CRM system

There are 3 primary ways the data in your CRM degrades in quality

1. Marketing enters incorrect, incomplete, or denormalized data, or corrupts good data
2. Sales people enter in duplicate, invalid, or incomplete data
3. The data just naturally decay in quality over time

The first source listed above is now largely fixable and is the subject of this post. The second is controllable with reports that find the issues, and management support to cajole the sales people into better behaviors. The third is fixable with regular data appends and cleanup.

There is no excuse anymore for marketing in the entering of bad or incomplete data. Marketing Automation Systems have the capability to validate and standardize data, rejecting data that does not measure up.  So what are the functions of a good response management funnel as it relates to maintaining data quality standards? If you have simple forms that use a web2lead like interface to jam data straight into your CRM, then you have a highway to data decay.

Here is a 90 second video on the Market2Lead Response Management process. Also here is a good second video, 3.5 minutes in length, on all the data validation and rejection features for keeping your data clean.

To maximize data quality, and richness over time, you will want your on-form capabilities to include:

• The combination of City, State, and Postal code is validated
• Email format validation (has to have a “@” and a “.” for instance)
• Phone number format validation
• Postal code format validation
• Rejection of common junk (asdf, xxx, abc, qwert, etc) in names
• Rejection of public domain emails for high value business to business offers

Once the form is accepted and submitted, there is still plenty to do. You are likely to de-duplicate with your database on email or some combination of name and company name and components of the address.  Only then do you send them to the fulfillment page (or confirmation page).

You still have more processing on the response however.  At Market2Lead we do another round of rejection, this time against a list of 600+ expletives and other junk words people enter into the name fields. The interesting part here is that this record is still sent into the database, but with a status of “Rejected”. The reason you want to  send “rejected” responses to the database is that it allows the marketer to identify common sources of junk submissions and also check for “false positives”. The rejected submissions are easily deleted at any time because of the “rejected” status.

The next step is augmentation. This is a fantastic idea, and highly undervalued. You can use it to do all sorts of stuff like assigning sales territories to leads, or converting state abbreviations into full length state names.  You could create a flag to indicate if the person used a public domain in their email address, or automatically tag registrants as competitors based on their email domain. You can recognize new leads from existing customers and flag them so they are sent to the account owner without delay.

After all the data rejection, augmentation, and suppression is done, then it is time to score the leads, and decide where to route them.  With strict data standards, documented process, and automation to prevent bad data from getting into the system, your marketing campaigns will be more successful, and your sales people will be more appreciative of the data they get from marketing.

-Kevin

Data quality, Customer Data Integration, and the Hippocratic Oath

A few months after my dad died back in 1984 my mother received a letter from our bank addressed to “Dr. T.B. Joyce (deceased)”. Needless to say my mum was a little distraught.

In a similar case I learned recently that the sales people of a local company were putting the gender of the person in parenthesis after the person’s androgynous first name. Imagine getting a personalized email from that company:
Dear Terri (female), or Dear Pat (male)

In a post back in March I discussed the many ways that data goes bad. Excluding the slow natural decay of data (actually it has sped up in the recession), sales people are responsible for much of the bad data. In the past Marketing has also shared the blame but the marketing automation systems now have the ability to validate incoming data, do format checking, and generally minimize the likelihood that marketing is adding bad data to the CRM. To wit, it is time that Marketers took the Hippocratic Oath to vow to do no harm to the Customer Relationship Management database!

The Hippocratic Oath for Marketers
I SWEAR in the presence of the CRM Administrators and before my marketing peers that according to my ability and judgment I will keep this Oath and Stipulation.

I WILL CONTINUE with diligence to keep abreast of advances in marketing. I will import only valid, well formed data, with all required fields, into the marketing automation system, so long as the treatment of other data is not compromised thereby, and I will seek the counsel of particularly skilled IT dudes where indicated for the benefit of the database.

I WILL FOLLOW that method of data treatment which according to my ability and judgment, I consider for the benefit of the database and abstain from whatever is harmful or mischievous. I will neither prescribe nor administer a lethal dose of dubious updates to any database even if asked by my VP.

WITH PURITY, HOLINESS AND BENEFICENCE I will pass my life and practice my art. Except for the prudent correction of an imminent CANSPAM violation, I will neither wholesale overwrite CRM data nor carry out any new field experiments on the database without the valid informed consent of the CRM administrator or the appropriate Sales Ops designee thereof, understanding that field insertion must have as its purpose the furtherance of the health of the database. Into whatever database setting I enter, I will go for the benefit of the data and will abstain from every voluntary act of data corruption and further from the addition of duplicates.

WHATEVER I MAY SEE in connection with Marketing driven CRM data changes which ought not be spoken abroad, I will immediately divulge, reckoning that all such should not be kept secret from Sales Operations.

WHILE I CONTINUE to keep this Oath unviolated may it be granted to me to enjoy life and the practice of the art and science of marketing with the blessing of the CRM administrators and respected by my peers and society, but should I trespass and violate this Oath, may the reverse by my lot.

My dad would be proud!  Got your own data horror story, send it my way!
-Kevin

How did my data quality get so bad?

As a general rule in marketing we believe that static data decays in quality at the rate of 1% per week.

Why is this important? Consider the CANSPAM laws. Your Salesforce.com account allows you to have multiple copies of the same person based on their email. Your Inside Sales Rep flags one of the copies as an “unsubscribe” at the request of the individual. Are you sure that is the same contact record your email vendor is using? Or consider the budget you are wasting by marketing to a database that you know is riddled with data problems, duplicates, invalid data, and then you get nailed for having lower open rates and click-through rates than the industry average! It’s the old 60:20:20 rule. 60% of the success of a marketing campaign is on the quality of the list, 20% on the quality of the offer, and 20% on the packaging. Good data is the keystone of successful marketing campaigns that bridge the gap between you and your prospect.

Good information quality will allow you to get the right data, to the right people, in the right place, at the right time, in the right form, at the right cost, so they can make the best business decisions. There is one thing worse than a lack of information, and that is bad information, because the latter can lead to bad decisions, whereas the former will generally stimulate more consideration before making a decision. Marketing Automation Systems can certainly help ensure that the data flowing into the system is much higher quality and does no more harm. But there are multiple ways that data can decline in accuracy and result in poor information quality. The following examples will help explain:

• Information quality decay (gradual loss of accuracy, loss of consistency)
  • People and companies change their email, addresses, phone numbers, names
  • Fortunes change, peoples titles change, and corporation sizes vary with time
  • Someone renames a campaign, or changes filters in a report and the trend data no longer has a common referential basis
• Data that becomes ‘corrupted’ by being overwritten with incorrect data (loss of integrity)
  • Marketing fills in a demographic field, followed by a sales person revising with improved data, followed by Client Services changing it back to original value
  • User with incorrect permissions accidentally changes a record
• Data that is a duplicate and is therefore double counted in reports (loss of cleanliness)
  • Sales rep adds new account or contact because they didn’t find it in the system
  • Marketing imports leads from show and accidentally adds duplicates
  • Prospect uses two different email addresses
  • Syndicated content provider uses Salesforce.com Web2Lead interface and introduces duplicates
• New data added that is just plain wrong  (loss of accuracy, lack of completeness)
  • Leads entered directly by Sales may not go through the field validation rules that marketing has access to with the Market2Lead Marketing Automation System
  • Auto-spammers fill in forms on website (increasing problem)
  • Prospect mistypes their own email address
  • Incomplete fields (the absence of some required members of a data set)
  • No common defined referential set for data exchange between ERP system and Salesforce.com  results in data being uploaded to the wrong field.
  • Person entering the data gets lazy and picks random values from pick lists for things like industry and role
  • Form attempts to collect BANT data far too early in the relationship with a prospect and as a result the prospect enters bogus data.

Over the coming months I plan to blog more about ways to tackle the data quality issue, especially if you have Salesforce.com.  Dealing with bad data requires a mix of process, people and technology. The activities can be divided into those that prevent data decay, and those that detect data quality issues and subsequently help remediate those issues.  The Market2Lead Marketing Automation Technology has a terrific response management engine that validates addresses, rejects junk like asdf.com, suppresses public domain email addresses, normalizes data, and ensures all required fields have data in the required format. But you still need to work on the People and Process to get really get a handle on data quality because your Sales Reps and Client Services folks still get to put data directly into Salesforce.com and  Garbage In means Garbage Out (GIGO).

That's all for now, more on detection and remediation later, Kevin

7 Best Practices to ensure Sales and Marketing Funnel Integrity

Last week we launched a new Demand Generation campaign that ran into a minor hiccup prior to launch. None of the Leads we were capturing were getting published to Salesforce.com. Upon troubleshooting we found out that there was a data type mismatch between a field we were capturing in marketing and the corresponding field in Salesforce.com. Our marketing process was asking for a range of values for Annual Revenue:

• $1 million to $50 million
• $51 million to $250 million
• $251 million to $500 million
• $501 million to $1 Billion
• > $1 Billion

But Salesforce.com expects Annual Revenue to be a number field on the Lead object in Salesforce.com. One of our new operations team members made this change without knowing the constraint on salesforce.com. Fortunately, our system allows us to have one set of values that a Prospect sees on a Landing page and another set of values that can be published to Salesforce.com. We were able to quickly fix the problem and launch the campaign.

This is one of the many process and data related issues that can break the integration between marketing and sales funnels. Here is a list of best practices that we have built over time to ensure that the integrity of marketing and sales funnels is always maintained.

Guaranteed Delivery - Ensure mandatory information is always captured: Most SFA systems require a few fields to be always filled in. For example, in Salesforce.com the default configuration expects Last Name and Company Name to be required fields. Most companies make modifications to this default setting and might add more fields to the list. Some typical fields I have seen are Email, Work phone, Job Title, Time frame, Budget among others. Marketing processes might or might not be able to collect all the required information from online and offline interactions. In such cases ensure that there is always a default value that the system can assign to these fields. For example, at Market2Lead we assign a default value of 'noemail@market2lead.com' for email, 'UNKNOWN' for Budget. This will ensure that Leads are propagated to the SFA process.

Appropriate Assignment - Ensure Lead's get assigned to the right sales rep: The one process misalignment that I am very cautious of is broken lead assignment rules. I have seen a few instances where customers find hot leads that require timely follow-up are stuck in some administrator's queue because the lead assignment rules were not triggered appropriately. Marketing processes should ensure that data values that trigger lead assignment rules are always in synchronization between marketing and sales systems. If the sales process uses geo based routing then check whether the rules are based on ISO codes or full names for states and countries. If the routing is more complex like a combination of geo information and product mix then ensure that along with geo values the product names match between the sales and marketing systems. At the end of the day administrators of marketing and sales systems need to keep their systems in sync to ensure Leads end up in the right rep's inbox.

Provide Progressive Intelligence - Search before creating a new Lead: One of the goal's of Lead Nurturing is to keep the sales reps' abreast of marketing activities and their Lead's response to these activities. This objective is defeated if the marketing system is not incrementally updating the Lead record that the sales rep is looking at and instead creates a new record for every interaction. Marketing systems should always search for an existing Lead in the sales system before creating a new Lead. If a match is found the existing Lead should be updated. The search and merge criteria has to be robust too. Here are a few things to keep in mind:
    a. Email is not a water tight method to detect existing Leads. In fact, reps rely on phone numbers more than emails. So, the detection process should be sophisticated to include a combination of multiple checks like email, first name + last name + company name, first name + last name + work phone among others.
    b. *NEVER* update key fields in the sales process without a rep's permission. Marketing systems are good at collective basic information through online interactions but are not always the most reliable method to capture critical fields like phone number. Imagine the plight of a sales rep if you update a real phone number with a bogus number like 111-555-1212 that marketing collects from an online form. marketing should work with the sales team and identify fields that can be automatically updated and fields that require a sales rep's manual oversight. In the second case it's a good idea to publish new information to the comments field and allow the rep to make appropriate changes.      

Align with Sales Follow-up Process - Make it easy for reps to access marketing Leads: Sales reps like to spend as much time as possible selling (and rightfully so). Therefore, it is the responsibility of the operations team to support the sales team by making it easy for them to access Leads from SFA systems and also minimize any change to the way they are used to interacting with these systems. Marketing might be providing top quality Leads to the sales but will not be assured of appropriate follow up if they make it difficult for the reps to access these Leads. Before setting up an automated Lead
delivery method marketing teams should familiarize themselves with the method/methods sales reps use to access Leads from the sales system. In Salesforce.com I have seen different methods being used. A couple of examples are:
      a. Rep based views: Sales rep's have their own views setup and they look for new Leads within this view. In some cases the filter criteria for the views is standardized across the organization and in other cases it varies by territory, product line or other criteria. Get a holistic view of how these view's are structured.
     b. Dashboards with alerts: In other cases I have seen sales organizations use more sophisticated dashboards that have a combined view of Leads, events, tasks, alerts and Initiative based follow-up procedures. Get an understanding of how these dashboards are setup. 

Ensure Timely Follow-up - Ensure sales reps are sent alerts when Leads respond: It is always a good idea to alert the sales team when marketing is able to generate a response from an existing Lead or generate a new Lead that requires timely follow-up. Get a good understanding of how the sales team is setup before coming up with an alerting scheme. For example, in some organizations all follow-ups triggered by marketing have to through an Inside sales organization. In other cases the follow-up process is dependent on where the Lead is in the sales cycle. Early stage Leads will require a follow up by inside sales and advanced stage Leads/Opportunities require follow up by field sales. Depending on the need alerts can be delivered to the the rep's dashboard in the SFA system or can be a combination of dashboard alerts and emails to the reps. Ensure you strike a balance between alerting the sales team and bugging them :-).

Avoid Data Black Holes - Align data standards between marketing and sales systems: I have seen quite a few companies suffer from what I call the 'Phantom Record' syndrome. One will be surprised how many records get buried in the database and don't show up on anybody's dashboard purely because the data values are not visible in the UI. CRM systems are notorious for being very forgiving when it comes to enforcing data integrity checks. For example, salesforce.com allows data to be imported to a pick list that does not match the pre-defined list of values. Here is a typical example. Let's say, a company uses personalized Lead views to allow their sales team access Leads and these view are geography based. One of the area manager Joe's territory includes California, Nevada and Oregon. The view is based on 2 character ISO codes CA, OR and NV. marketing publishes a set of hot leads from California and the state name is set to 'California'.  Joe will never see these Leads since the view will not pull these records and Joe can't access them manually too since the state pull down menu in the UI has 2 character codes. It is always a good idea to have proactive oversight procedures in place to avoid users violating data policies when systems can't enforce them. The next bullet talks about reactive procedures that can be put in place to catch issues that slip through the cracks.

Track Outliers - Setup Exception reports to track outliers: It might not be feasible to systematize all the processes and best practices and build logic  based rules to catch errors pro-actively. As the saying goes, it's better to be safe than be sorry. Always augment proactive checks with reactive exception reports to catch outliers. Some typical reports could be:

    a. Weekly lead flow reports: A mature marketing and sales process will provide guidance into typical lead flow on a weekly basis. Setup a 10%-20% tolerance and trigger alerts if the system detects unusual Lead flow. A few month's back we were working with a customer who used to get regular data feeds from an external source. At some point one of the system parameters (FTP location information) was changed by the operations team to consolidate data feeds from multiple sources. The receiving function was not aware of the change and for a few weeks these Leads went untouched. The problem was caught and rectified before too much damage was caused. An exception report would have caught a sudden decline in the Lead flow and the resolution could have been much more timely.

    b. Weekly lead follow-up reports: It is a good idea to track how many Leads are being followed up by sales. This is a good way for marketing to stay on top of any process changes that might take place within the sales team that requires a change in the Lead publish method. For example, in a particular situation, one of our customer's changed the 'Lead Status' values and the associated follow-up business rules. The marketing team was not made aware of the change and the workflow rules were still setting 'Lead Status' values that was no longer used by sales. With an exception report the marketing team realized that existing Leads were not being followed up on and upon investigation found the cause and made appropriate changes. In this case the exception reporting process was in effect and helped catch the problem right away.

To conclude integrating marketing and sales funnels requires proper attention from both marketing and sales teams since it's a cross functional responsibility. Over and above the 7 steps I mentioned above, a robust communication protocol is extremely critical to ensure that the two functions are always working in tandem. This might sound banal but I still get surprised to see many organizations not investing enough in this critical integration that can directly impact top line revenue.