Measurement and Metrics


After a successful first half of her day and a good break, Janet is ready to tackle the tasks ahead of her.  At 1 pm, Janet reviews the latest campaign metrics.  She notices that one of her campaigns is not getting good response rates.  She calls the campaign manager into her office to brainstorm getting this campaign back on track.  It is determined that the messaging and offers should be tweaked.  The campaign manager goes off to make the changes and they agree to discuss the results in a few days.

Next she takes a look at the lead scoring results.  She notices many leads in “hot” status.  A quick look in the CRM shows that many of these leads have not been followed-up on by the lead qualification team.  She heads over to talk to the lead of that team.  The conversation is easy because the teams are in close alignment on strategy and incentives.  Everyone is working towards the same goals.  She determines that the issue is around having too many leads in hot status, but when the lead qual team follows up, not many are converting.  She discusses with her Marketing Automation administrator some changes she wants to make to the lead scoring criteria.  The changes will be made by the end of the day and they can review the results in a few days again to see if it helps.

At 3 pm, Janet interviews a couple of candidates for her team.  She is looking for a business analyst who can help with reporting and data quality.  This will be a key role to keep her team supplied with the right information and prospect data to be effective in their campaign efforts.  She is looking for someone who can not only crunch the numbers but also analyze the results and bring suggestions to the table.  She also wants someone who is familiar with data best practices that can keep their database clean and complete.

At the end of the day, Janet looks back at her dashboard to find some campaigns that are working well.  She shoots off some meeting invitations to business leaders to discuss how she might be able to replicate these campaigns with their content.  She updates the campaign calendar based on her earlier strategy discussions and shuts down her computer for the night.  She’s now off to watch her son’s soccer game secure in the knowledge that her marketing department is an effective and efficient piece of the company’s revenue engine.

Janet gets to the office around 7 am because she likes to work early to avoid interruptions and be done with work at a reasonable time so she can spend more time with her family.  However, her CFO feels the same way, so this morning she is accosted with questions as soon as she walks in about the effectiveness of her marketing spend.  Fortunately, Janet has this information readily on hand because she has a marketing dashboard that gives her real-time information about closed-loop marketing ROI.  She can tell the CFO exactly how much she has spent and what she has in both closed won and anticipated revenue for the month, quarter and year.

At 9 am, Janet has a meeting scheduled with her CMO to talk about the marketing strategy for the next quarter.  She walks in with information about what has been working for them and what hasn’t.  She also has some ideas about what they could try because she has been quietly testing out new strategies for the past few months and has information about what might work.  The meeting concludes promptly at 9:30 with a clear vision of the strategy for the next quarter and buy-in from the CMO.

At 10 am, Janet chairs a staff meeting where the Revenue Marketing team reports out on the status of current campaign projects.  With simple workflow tools and project dashboards, Janet can clearly see where everything stands and get tasks at risk of slipping back on track.  Everyone on her team has clear roles and understands how they are evaluating campaigns against revenue.

At 11 am, Janet sits down to review the latest content pieces the team has developed.  She has clearly defined personas and buy cycles for her prospects, so she knows what content will be effective in each stage of the decision process for each type of prospect.  That makes her job in editing the content easy.

At noon, she finally gets a break before gearing up for the second half of her day…

If you knew in advance that your campaign would bomb, you’d never run it, right?  If you are a traditional marketer, you probably know how to evaluate which publications to run an ad in or whether an event is useful to attend.  And you evaluate these things up front before ever investing the resources or money in them.  The same should be true of your lead generation campaigns.

The difference with lead generation campaigns is that you should have a much more specific goal in mind, such as generating a certain number of leads or revenue from the campaign.  The best way to figure this out is by modeling based on past history.  If you’ve done campaigns long enough, you should know your average conversion rates.  But even if you don’t have past history to draw from, there are enough baseline metrics out there for B2B firms, you should be able to find a reasonable set of metrics to model your campaign.

At it’s simplest, here’s what a model might look like to determine the expected revenue of a campaign:

# Inquiries x % Conversion to Opportunities x % Close Rate x Average Deal Size = Expected Revenue

You may also need to add in information about channel performance and number of targets to figure out how many inquiries to expect.  If you tend to market to a lead throughout the sales cycle, these numbers will generally be reflected in your conversion rates, so you are really looking for number of inquiries that this specific campaign might generate.  If this campaign is meant to be one that pushes people through the pipeline, maybe you are more concerned about increasing the conversion rate or speeding cycle time.  Even if you don’t have numbers, you can model out what it might look like if you increase the conversion rate from a marketing qualified lead to a sales qualified lead by even 1%.

If you don’t have a past history, at least create a model with some conservative numbers to track against.  Once you do this a couple of times, you will have that history, but you need to see if your assumptions are reasonable.  So start making some assumptions to test.  Bottom line – never run a campaign at all before you at least take the time to see if it can theoretically meet your goals.

Customer Relationship Management (CRM) implementation projects are generally big hairy beasts that take a long time.  Especially if you are a large global firm with lots of business units that need to cooperate to figure out a data structure that works.  So it’s not surprising that the marketing department, tiring of waiting for all the business units to duke it out on CRM, is often tempted to go around them and setup a Marketing Automation (MA) system on their own.  Most MA vendors have a cloud-based option for their software today, so marketing can get this done without IT and without sales, which means they can get it up and running very quickly.  But there are a few issues with this that you need to be aware of up front so you can address them.

Problem #1: MA systems do not manage leads.
MA systems are meant for the marketing department to use.  They have the ability to score and route leads, nurture leads, and notify people of leads.  But leads don’t live in MA systems.  MA systems track contacts and activity on contacts.  When there is sufficient activity to call a contact a “lead,” that lead needs a new place to live where the sales team can view and track its progress and ideally turn it into an opportunity.  Salespeople don’t log into your MA system, nor do you want them to.  Even if they did, there is no easy place for them to find their leads.

Problem #2: MA systems do not track campaigns.
The second issue is that MA systems don’t have a strong concept of campaigns.  They have silos of activities and programs.  But a true concept of a campaign with several parts that can easily be tracked together with a full lifecycle of contacts, leads, and closed won sales just isn’t there.  Some MA systems have tried to implement pieces of this but with little success.  The MA system is really meant to be the execution engine, not the tracking mechanism.

Problem #3: MA reporting is limited.
MA systems are going to report on contacts and activity.  It will not let you easily report on lead progress, revenue segmented by your critical criteria and multi-channel campaigns.  Some MA systems have tried to incorporate some of these pieces, however I have not yet seen anything that comes close to the type of reporting that an enterprise firm needs.

Problem #4: Data has to be refreshed manually.
Another issue is that your data will only be updated when either you upload new data, or your customer fills out a form.  Your salespeople will not be able to update contact information.  You likely also have many other systems and departments, such as accounting, that might have updated data.  All of this information would have to be ported manually into the MA system and data priority would need to be managed manually.

So now that you know the problems, what are the possible solutions?  Integrating a CRM to your MA system is obviously the best choice.  To get robust reporting and segmentation, you should also integrate your other systems (ideally into your CRM or data warehouse as the data master).  If this just isn’t an option or your CRM implementation is taking a painfully long time, there are plug-ins out there that you can setup to temporarily manage your leads and reporting.  Just keep in mind that you will still have some of these limitations until you get your CRM up and running.

In continuing my post from last week, I wanted to share a recap of a couple of the other presentations from the @Marketo Revenue Rockstar tour in Chicago.  The first is from TPG’s own Debbie Qaqish (a.k.a. Lady Qaqish or @revenuemarketer).  I’ve seen Debbie speak many times on these topics but she’s always so engaging I never get tired of hearing her.

Debbie shared some insights on what it takes to become a revenue marketer.  She outlined a journey from traditional to lead generation to demand generation to revenue marketing.  There is a significant jump from lead generation to demand generation that takes the focus from leads and metrics like email opens to pulling leads through the sales process and metrics such as conversion rates and days to close.  You really have to have a marketing automation system with CRM in place to make this leap.  Then the jump from demand generation to revenue marketer is really characterized by making your process predictable, repeatable and sustainable.

According to a 2010 CSO Insights survey, sales are still having to generate most of their own leads and yet sales effectiveness is the #1 initiative for the VP of sales for several years running.  Which means most companies out there aren’t even at that demand generation stage.  Sales is crying out for marketing help and marketing needs to listen and respond.

The next presentation was from Ron Ens from the Lenskold Group.  As I’ve said in previous posts, I love numbers, so this was just a fascinating presentation for me.  The Lenskold Group helps their clients really dig into the metrics and measurement of marketing efforts.  He describes the maturity of measuring as a journey from tactical management, to lead quality, to strategic, to revenue and ROI.  You may think you are measuring revenue today, but trust me you can get much more sophisticated!

One of my favorite graphs Ron showed was a case study of a company who was looking at lead source (by last touch) and thought that their most effective tactic.  Yet when they actually looked at lead source through a shared attribution model, they saw that really direct marketing made a huge impact.  This is really a good example of that lead quality level of measurement – much more than just tactic by tactic.

At the strategic level, Ron’s team helps clients start forecasting revenue and gain insights on the right marketing mix and targeting/segmentation through either marketing mix modeling (using 2 years of previous data) or marketing testing (forward-looking).  And finally, at the revenue and ROI level, you really start to tie financial scenario planning into marketing – understanding both revenue and profitability (as Marketo is doing in their marketing efforts as noted in the case study in my last post).

Measurement, like the rest of your marketing efforts, should be a journey.  As Jon Miller notes, start small, think big and adapt quickly.

Hope this recap was helpful and inspiring.  Try to attend one of these great events or if you can’t there is also a live stream of the New York event on June 7th.

The imagery of a funnel is often used to describe the process that targets go through for marketing and sales.  But it’s not very accurate.  A funnel is technically used to control the flow of something.  Eventually, everything that goes into the top of the funnel will flow out the bottom.  While I applaud the optimism that every target will become a sale, that’s just not realistic.

A better analogy would be a sieve.  Or a funnel that someone poked a lot of holes in.  What happens in reality is this – you send out a marketing campaign and you get some people to respond.  Those people are going to be at varying stages in the buying cycle, or maybe not even in the buying cycle at all.  They may be at an early stage or they may just be browsing around thinking about future products and services.  Or just keeping up on what’s going on in the marketplace.  They may be your competitors or students.  There are going to be lots of people in that first group who you just don’t want to waste your time on.  So your next marketing materials or sales interaction will ideally start to weed out the people who are going to be good prospects vs. those that aren’t.  You get a little leakage out of the funnel or sieve.  Each subsequent conversation should weed out a few more of your original leads until you get down to the very best targets who are highly likely to purchase your product or service.

If you don’t have a leaky funnel, you probably aren’t doing a good job at narrowing in on the best targets for your sales team.  Your sales team has very limited time to interact with your targets, so you want them spending their time on the very best ones that are likely to buy.  If you are just sending them every lead you get, they are probably spending way too much time chasing bad leads and are likely to resent you in the process.  Your job in marketing is to make sales more efficient and help pinpoint the right leads for them to go after.

You won’t always get it right.  Sometimes you will filter out someone who will go buy from your competitor and sometimes you will keep someone in the funnel who shouldn’t be there.  But that’s okay – this is an iterative process.  By doing this over and over again and measuring every step of the way, you’ll figure out how to improve your funnel and just where to poke the holes to make it work well.

So next time you get a conversion rate of less than 100% from one stage of your funnel to the next, applaud your success in helping focus in on the right leads and just keep measuring everything.

My CMO used to constantly question the validity of my marketing tracking.  He would question how we knew that marketing made a difference and that the sale wouldn’t have happened anyway.  He would ask how we knew it was this event or that campaign that specifically led to the sale.  I would pour over report after report trying to sort this out in fear of my next meeting with him.  Was he right to be asking these questions?  Absolutely.  Did I ever come up with a great answer for him?  Absolutely not.

When you work in B2B services selling big ticket items that take 6-24 months to close, you just can’t find clean metrics that say this campaign brought in $X directly and that campaign didn’t work at all.  It’s all about aggregation.  If you look at activity and behavior over time, the picture can become clear.  One metric I found that I really loved was comparing marketing-influenced versus non-marketing-influenced sales.  You can look at the average deal size and time to close.  I found that on average marketing-influenced deals were larger and closed quicker.  That can’t just be coincidental when you are talking about large enough sample sizes.

It’s still worth it to track the source of leads, though, as long as you are clear about it.  Companies will generally tag the resulting opportunity with either the first campaign that created the lead, or the most recent one that created the opportunity (or both).  That’s still a meaningful metric if you are tracking that consistently – if certain campaigns are never generating leads or pushing leads over the edge, it’s at least worth reclassifying those as branding instead of demand gen campaigns.  You also need to be patient and track over long periods of time before you start analyzing this data.  A few months of data isn’t enough to decide a campaign was successful or not – you may have to wait a year before you do the final analysis.

My advice is this – track everything.  Track the original source of the leads, track the campaigns that push your leads into opportunities, track behavior.  Once you’ve done this for a while (and that may mean 6-24 months depending on your sales cycle), analyze it by going through some deals that closed in detail – look at the campaigns that were run, the contacts’ responses to those campaigns and the results.  Looking at five to ten accounts in detail will probably give you the information you need to figure out what is going on and what is meaningful.  Then you can narrow what you track and the metrics you focus on to report to your senior team.  While it’s not ideal, “I just don’t know yet” may have to be your stock answer for a while until you get enough numbers to make your reports and metrics meaningful.

I love reports.  You just can’t beat some good marketing metrics when you walk into a management meeting.  Unfortunately, one of the biggest challenges to measurement that I often see when a company is implementing marketing automation (MA) for the first time is not having any benchmarks.  If you don’t have any benchmarks, how do you know that you’ve improved your metrics?  As I talked about in my blog post from a few days ago, you will soon have massive amounts of data flowing in, so you need to figure out now how to measure before you get too far behind.

I have two suggestions for you.  The first is just to start measuring.  Pick some metrics that are meaningful and start measuring them.  Your measurement today is your first benchmark, and you can look back over time to see the trend.  Be sure you track these measurements outside of your MA or CRM system if your system does not “snapshot” data (i.e., most systems are just cumulatively adding data, so sometimes it can sometimes be hard to isolate certain metrics and trend them over time).  If you are having trouble picking your metrics, go back to your marketing goals.  If your goal is to make money through marketing, well then revenue related to marketing campaigns would be one of your metrics.

My second suggestion is to use industry benchmarks.  Here are some useful ones:

  • The average B2B firm spends 4-5% of revenue on marketing
  • Based on the Sirius Decisions model, about 13% of Inquiries will turn into Closed Won Opportunities
  • Also based on Sirius Decisions, about 23% of Sales Qualified Leads will turn into Closed Won Opportunities
  • The average open rate for a B2B email is around 20% and the average clickthrough is about 4-5%
  • For a form on a webpage, the average conversion rate is around 0.2%

You can probably find others that are more specific to your vertical – there is a ton of research that exists out there by Sirius Decisions, Forrester, and others.  A quick Google search will usually find you something you can start with.  Good luck and happy number crunching!

Any large company who has ever implemented a marketing automation (MA) system knows that you go from 0 to 60 in about 2 seconds in terms of the amount of data you start collecting on your clients and prospects.  It’s overwhelming.  And before you know it, you are completely paralyzed by all the data – it becomes meaningless to you.  That’s about the point you start wondering if you just wasted a ton of money on a really expensive version of an email marketing tool.

Well, the good news is there are a few things you can do to avoid the analysis paralysis and start making sense of all this data.  The first thing you can do is to take advantage of the automation part of that marketing automation tool you have.  One way to automate data is through lead scoring.  Figure out what those key pages on your site are or some key downloads you have.  When people visit those pages or download those assets, use that as an input to your lead score to figure out when they are ready for the next step in your sales cycle.  Another way to automate is to put people into campaigns or nurture programs based on what they are doing.  Think of the “if you like this, you might try…” feature you often see on B2C websites.

Another tactic you can try is limiting the metrics you look at, but measuring and looking at them consistently.  We challenge our clients to start out by picking the five best metrics that represent their goals.  For example, if one of your marketing goals is to generate better quality leads for the sales team, you might measure conversion rates of your leads through the sales funnel to see if they improve.  Most marketing automation and CRM systems will let you setup dashboards to keep track of these metrics easily.

A slightly more drastic measure, but one that you will find yourself in need of doing eventually, is setting up a business intelligence tool.  I have yet to find a MA tool or CRM that is capable of handling the complex matrixed reporting that large global companies tend to require.  At some point, although it might not be right out of the gate when you implement your MA or CRM, you will find that you need to move up to a BI tool to set up the reports you need to monitor the mass amount of data you are collecting.