Our Customer Service folks are working with a large client to implement a Sirius Decisions style dashboard for their leads. It shows the totals for new inquiries, marketing qualified leads (MQL), sales accepted leads, and sales qualified leads in any given time period. One of the trickier parts though is in documenting the conversion rate between these lead states. If you thought the answer was just to divide the converted leads by the new leads in the period, you might want to think again. It will be in your favor to decide on the best method to calculate these conversion numbers before you present the lead waterfall chart. Knowing the conversion rates gives you a strong idea on how many leads you have to generate to fill the sales funnel.
Calculating the lead conversion rate
Let’s go through the process for just a single lead status transition – we want to calculate the conversion rate for leads with a lead status of “Inquiry” to leads with a lead status of “Marketing Qualified Lead”, or MQL, in a given time period, let’s say first quarter – Q1. This should be a nice measure of what percentage of raw leads Marketing is able to qualify and hand off to Sales! Typically the answer is 10% - 20% for B2B.
Jan 1, there are 1000 leads sitting in the CRM database with a lead status (state) of “Inquiry”. By Mar 31, 300 new ones have been added to the database, and from the combined total of 1300, only 50 converted to “Marketing Qualified Lead”. i.e. their lead status value was changed from Inquiry to MQL! Of the 50 that converted to MQLs in Q1, 25 of them were from new inquiries created in Q1 and 25 were from the 1000 fossils that were sitting around when we entered the quarter on Jan 1.
So which of the following is the most useful or accurate conversion rate:
- 50/1300 – 3.8%
- 50/300 – 16.6%
- 50/1000 – 5.0%
- 25/300 – 8.3%
- 50/(all the Inquiries created in the last n days)
50/1300 – 3.8% Inquiry to MQL Conversion Rate
This may actually be the most “accurate” answer, but it is certainly
deceptive, and won’t make you a hero in marketing. The problem is this. If a
lot of leads come in, create one response, and never engage with you again,
they may remain in the “Inquiry” state for years. They never engage enough to
become MQLs. Since the “1300” number contains the 300 new AND the 1000 fossils
that were lying around when we entered the quarter, the fossil number could
continue to grow and grow. Ie your conversion number would shrink every quarter
even if you were converting new leads at the same rate! So the 50/1300 without some form of
expiration date for the fossils is a bad idea. More on the expiration later!
50/300 –16.6% Inquiry to MQL Conversion Rate
‘Number out’, in the
period divided by ‘number in’, in the period. It feels like the right thing to
do but… In reality only 25, not 50, of the 300 new inquiries converted so it is
deceptive, and if you use this method, and next quarter you do a fantastic job
creating leads and add 600 new inquiries, and the same percentage of the new
ones convert (50) and you get another 25 from the fossils, the conversion rate
is 75/600 - 12.5% that’s a lower percentage than 50/300….but you actually
did better because you generated more new leads! Definitely don’t want to use
this one!
50/1000 – 5% Inquiry to MQL Conversion Rate
This is the total converts in the period divided by the
starting number, and the conversion rate looks bad, and the more fossils that
gather in your database, the worse this number starts to look. This number is
the least likely to give people a gut feel for the waterfall. Don’t want this
one!
25/300 – 8.3% Inquiry to MQL Conversion Rate
This is the total of the new inquiries in the period that
also converted in the same period, divided by total new inquiries in the
period. This is very attractive in its simplicity but it is also dangerous. It
ignores all of the conversions from the fossils. If the period of measurement is
a year then the conversion number may be quite accurate, but if the period is a
week, it could be wildly inaccurate. If the period is a quarter, and most
inquiries convert to MQLs within 10 days it could be quite accurate, but if the
time to conversion for new inquiries typically averages closer to 60 days this
could be quite inaccurate. In our example, only half of the MQL leads in Q1 are
converted from new Inquiries in the period so this means of calculating the
conversion rate is only looking at half of the conversion data. The other
problem with this number is that it can be quite volatile. If you do a big lead
generation push in the last week of the period, and you don’t have a chance to
convert many of those leads to MQLs, the conversion rate could look bad that
quarter!
50/(all the Inquiries created in the last n days)
I like the idea of doing something like 50/total number of
leads that had a status of Inquiry in Q1, and are younger than twice the
average amount of time it takes for an inquiry to convert to an MQL. Let’s say
that in our example above the average lead, if it is ever going to convert to
an MQL, does so in 75 days. Chances are good then that after 150 days if an
Inquiry hasn’t converted it can be put out to pasture. So we could consider for
conversion purposes only those leads younger than 2x the average time to
convert, and look at the conversion rate of this young bunch. We ignore the
fossils.
So for our example above we would go back 150 days (2x75) from March 31, count all inquiries created since then that still had lead status=inquiry in the period Jan1-March 31, and use that as the divisor. If we created 400 new inquiries in Q4, but 25 of them converted in Q4 leaving 375, we would have a divisor of 300+375= 675. The Q1 conversion rate is 50/675 = 7.4%. So we are excluding all the truly old fossils from the 1300 number (total leads with lead status =”Inquiry” in the database in Q1), and looking at the younger pool of Inquiries that are most likely to convert. You will notice that this number is closer to the conversion rate for just the new inquiries (25/300=8.3%), but will be less likely to have the same volatility, and is more accurate. On occasion some of the old fossils (older than 2x the average time to convert) will convert so your "real" conversion rate is slightly higher than this. But let’s save that discussion for another long tail (sic).
Once you decide to pick a way of calculating conversion rates, stick with it, because the trend will become just as important as the raw number!
-Kevin
Comments