You lead a unique sales organization. You operate in a niche industry. Or your business model is evolving.
You are a data-driven decision maker, but your sales data makes problem solving difficult.
Your sales challenges can’t be solved by a one-size-fits-all analytical approach.
Let’s say you have 100 people on a sales team. Can they all be great? Most sales leaders say no. Conventional wisdom will say something like ten of them can be exceptional. Ten of them will be dogs. And the other eighty will fall somewhere between “solid performers” and “good enough to not be fired.”
I hate conventional wisdom. Especially when it’s as destructive as this. And make no mistake, this is destructive logic with little more than a baseless rule of thumb at its core. Nonetheless, most sales leaders pretty much accept the 10-80-10 performance standard as an incontestable law of nature. In so doing, we are inherently setting an artificially low bar with which to measure our success. Provided you have a great offering and a large enough target market, the goal of a sales leader should be to ensure that every salesperson on the team performs spectacularly.
Ironically, the KPIs – key performance indicators – that many sales leaders rely on to measure, predict and ensure sales rep success often actually perpetuate tiered success levels in which too many salespeople fail to meet their potential.
The problems with KPIs stem from sales leaders’ perceptions of KPIs and the opportunity cost of not taking a more scientific approach to understanding a company’s actual predictors of sales success:
Unreasonable Expectations & Emphasis
Think about your own sales KPIs. Do you have salespeople who hit all the KPIs, but still land on the lower half of the leaderboard? Do you have people at the top of the leaderboard who haven’t hit all their KPIs? Of course you do! In itself, that’s not a problem, but simply illustrative that most KPIs aren’t the locked and loaded predictors that sales leaders might wish they were.
KPIs are usually created with the right intentions. In some cases, they are meant to get out ahead of problems before they become systemic. In other cases they might be intended to set minimally acceptable levels of effort or output required by every member of the sales team. In these senses, KPIs can be directionally helpful predictors of success and failure.
However, KPIs are usually created from team averages without a detailed understanding of individual rep strengths or nuanced attributes of individual deals. Too many sales leaders overlook these differences and instead see the KPIs as black and white predictors of success. This becomes management by numbers and these leaders have teed themselves up for disappointing surprises.
When generic KPIs are overemphasized, they incentivize the worst possible behaviors at all levels of the sales organization. Sandbagging, or hiding deals, is heightened in organizations that overemphasize sales cycles or uniform stage velocities – even though everyone will acknowledge that there may be legitimate and compelling reasons for certain deals to progress more slowly than others. When organizations overemphasize metrics for forecast calls, reps often feel compelled to “commit” deals that they know are a little more than a shot in the dark. Organizations that lean too heavily on activity metrics often find reps logging questionable calls or sending superfluous emails that don’t bring a buyer any closer to a decision. Companies that inflexibly demand a 3-to-1 or 4-to-1 pipeline ratio often see leads prematurely being categorized as opportunities, and sacrifice quality for quantity.
The result is a morass of bad sales data that only pushes the target of understanding the true drivers of success and failure farther away.
Ignoring Half the Story
When defining new KPIs, sales leaders often look at attributes of the deals that won or the sales people consistently at the top of the leaderboard and attempt to draw conclusions. That seems like a reasonable starting point, right?
As a starting point, dissecting attributes of the best salespeople or winning deals is fine. But if the process ends here, sales leaders are positioning themselves for disappointing surprises because they haven’t considered the traits of underperforming reps and deals that lose. Too often, we find that companies’ chosen KPIs would have varied only negligibly had they been based on attributes of losing deals or reps outside of the top 10%.
What worked in sales yesterday, doesn’t always work today nor is it guaranteed to work tomorrow. Yet too many sales organizations are built around KPIs that haven’t been re-evaluated in years or don’t more heavily weight recent events.
And versus Or
Think about your company. What is fundamentally different about the deals that your organization wins from the deals that you lose? Is it certain buyer attributes? Is it certain seller attributes? Price? Product? Activity levels? Does sales cycle or stage velocity mean anything? Competition?
Most sales leaders intuitively recognize that all of these variables (and more) simultaneously impact every deal in the sales pipeline. Success and failure are not decided by one variable or another. They are the product of lots of different elements all moving simultaneously. Yet most KPIs are derived from single variable data silos, without any regard for nuances that might predict individual deal outcomes and rep success levels.
Great salespeople are dreamers. They naturally “think big” and have an “eat what you kill” mentality. They are smart, creative and competitive. Being less than spectacular is unfathomable. They have an ego, but they’re not arrogant. They will learn from the best, but they need to be better.
This is the DNA that we all hire for and it’s the DNA that positions us all for success.
When we forcibly jam these forces of nature into a sales mold based on perceived success drivers for a collection of other reps, we are discounting each individual salesperson’s individual strengths, weaknesses and abilities to succeed outside of that precise mold. Instead of capitalizing on the attributes that position each salesperson and each opportunity in the pipeline for success, we are creating cadence-driven robots instead of thoughtful business leaders.
Jim Dries is the CEO of piLYTIX, a data science as a service company serving heads of sales and their teams.
Every day of the week, it seems that a new “big data” company comes on the scene. Some of them are good and provide real benefit. Others prey on the people without a mathematical education and their fears of the unknown. Calling these guys pond scum or snake oil salesmen seems too cruel to pond scum and purveyors of even the finest snake oil.
At its very best, great data science can serve as a foundation for tremendous efficiencies. When it works, organizations can achieve greater sales and lower their costs. At its worst, when data science is misused, misunderstood or contorted into “management by numbers,” great confusion results and sales leaders’ abilities to impact results are lessened. So many organizations are struggling to define the proper roles that data science and analytics should play in their sales organizations. As such, it’s important to frame the proper roles of data science in our organizations:
Should: Help you make better decisions
Nearly every major decision made throughout a sales organization can be influenced by data science: Which opportunities and leads should get more attention or less attention. Where we should focus our sales, marketing or management resources. Which salespeople need more coaching and training. The variables that created the insight should be clearly understood, so that your colleagues have faith in the decisions you make.
Shouldn’t: Attempt to make those decisions for you
Data should influence decisions, but you were hired to be the final authority on those decisions. No statistical models will ever be driven by perfect data and no model can ever incorporate every conceivable data point. As such, a valuable data model should never produce a perfectly binary output telling you to absolutely do something or not. There is no silver bullet. There is no holy grail. An analytics person who tells you that there is only one way to accomplish your end goal is misinformed.
Should: Help you innovate
A scientific understanding of your team’s sales data will open your eyes to the areas in need of the greatest improvement. When the solutions to these challenges aren’t obvious, data science can empower sales leaders to effectively measure the results of small controlled experiments.
Shouldn’t: Be seen as a replacement for creativity
Data science isn’t a crystal ball that magically shows the future. At its core, it finds historical patterns of data – often very subtle patterns – and compares them to current data to give insights about most likely outcomes. It is not management by numbers. The best sales leaders use data science for the direction that it provides, but constantly look for creative approaches to motivating their sales teams, engaging their prospects, and scaling their processes.
Should: Empower your salespeople
The best salespeople are driven, smart and competitive. They need to win. They are constantly looking for advantages that will help them win the next sale, beat their competition and climb the leader board. Great data science can bring those advantages. It can help focus their limited time on those leads or opportunities that are most likely to win. It can help them remove their own blinders, and give them insights that will position them to play to their strengths and minimize their weaknesses.
Shouldn’t: Replace their humanity
Data science should never be conflated with a loss of humanity. There is no argument that should ever be accepted that would suggest that data science and traits that embody the best humans and the best salespeople – empathy, dignity, perseverance, intelligence, strength, humor – are somehow mutually exclusive. Data science can identify your best prospects. It can identify strengths and weaknesses in your deals. It can find very subtle weaknesses in your sales approach. It can even pinpoint causes for having bad sales data. Data science, however, cannot make an emotional connection to a prospect. It can’t carry a conversation, convey the value of your offering, or address competitive differences. Maybe one day robots will replace all of us, but they aren’t here today.
Jim Dries is the CEO of Austin, TX based piLYTIX – a Data Science as a Service company.
Sales enablement is suddenly very hot!
Just a few years ago, very few companies had a sales enablement leader. Almost no one had a full team. Now, most large B2B sales teams are investing heavily in sales enablement departments.
I talk to sales enablement professionals every day of the week, but I still struggle to define exactly what it is. I can’t define sales enablement because I am not a sales enablement practitioner and there is an awful lot of debate in the community of practitioners around what the role should entail. Like any nascent stage business discipline, there continues to be rapid evolution of the role.
I think that the word itself – enablement – is driving some of the confusion. The word is open to interpretation. And has it ever been interpreted differently by different people! When we google “sales enablement” we find scores of varying and complex definitions.
I debated this topic over breakfast with a founding member of the Sales Enablement Society a few weeks back. He has been an influential thought leader in this space for a long time. He opined that much of this debate could have been avoided if we had used a simpler term than enablement from the beginning. He suggested a term that gets to the heart of what everyone in this field is ultimately striving to achieve for their companies:
Now, when I go back and review all those complex definitions of the function, productivity seems like the perfect word.
But even if everyone agrees that productivity is the desired goal of the sales enablement function, there are still basic structural questions lacking consensus. These outstanding questions threaten to prevent the function from achieving its full potential and revolve around things like reporting structures and the level of interaction with other functions (for example, To whom should the enablement function report? What is the “correct” level of interaction with other departments like marketing and H.R.?). The question that my company often discusses with enablement professionals is the role that data science and analytics should play.
Some sales enablement leaders see data science and analytics at the core of their roles. Others, not so much.
If sales enablement is inherently about driving sales productivity, shouldn’t the sales enablement function be tasked with a very detailed understanding of the driving forces of success and failure for their companies? Why certain opportunities win and others lose? Why some salespeople successful and others decidedly not? Why certain goals are achieved and others are missed? What value sales technologies are bringing to the company? What content is needed to drive one sale or another?
The answers to these questions are never black and white. They can only be found through a scientific examination of a confluence of several different data points: buyer attributes, the salesperson’s strengths or weaknesses, content that was used, activities, technologies that were used, stage velocity and on and on.
To drive true productivity, the first order of business should be to have a thorough understanding of a company’s drivers of sales success and failure. As sales enablement practitioners continue to debate their roles and increase their influence within their organizations, data science must become a driving force.
Jim Dries is the CEO of piLYTIX, a sales enablement technology company.
Because we would rather talk about you.
Do any of the challenges at the top of the “About You” section speak to your needs? Is there something particularly unique about your sales organization that makes solving these challenges difficult? If the answers to these questions are ‘no,’ our products and services and services won’t be relevant to you. Nor would a one-size-fits-all pricing scheme. On the other hand…
If something you saw resonated, let’s talk.
Take a look at the “About You” section. If you can identify with a specific industry or specific sales enablement challenges, we might be a good mutual fit. Our team of mathematicians have designed a technical platform that helps our users address these issues.
Most of our users employ between 50 and 500 full time sales people who directly enter data into a CRM system. We have worked with companies as small as 10 salespeople and as large as 1,500.
However, we aren't for everyone. So that we don’t waste your time or ours, let’s also consider that we likely won’t fit with very early stage start-ups or companies that have just a small handful of sales reps. Likewise, we won’t likely add significant value to sales leaders who claim to have a perfect understanding of their perfectly maintained datasets.
We know of roughly 50 software companies who play in the loosely defined space of “sales analytics.” They are pretty easy to find, but if you’d like, we would be happy to send you contact information for all of them. We will let them define themselves and encourage you to reach out to them if you are interested.
We don’t spend a penny on competitive intelligence and wouldn't share our opinions of another company if we had any. We offer a smart product for smart business leaders and their teams. If you have found this page and feel that you meet our criteria, let’s talk about your needs and see if we can address your challenges. Worst case scenario, you will walk away from an initial conversation with us with lots of contact information. If you decide someone can address your sales challenges better than we can, we wish everyone good luck.
piLYTIX extracts data directly from your company’s CRM system or data warehouse. Our models examine fields that are native to the data source (account information, opportunity information and activities) as well as external technologies that have been integrated (lead scoring or proposal creation software, as examples).
It would be a dreadful idea to have a predictive model that presupposes any fields or combination of fields as statistical predictors of success (or failure). That would compel you to modify your data structure to our needs (which you don’t have time for) or the output would be of questionable value.
piLYTIX’s data scientists will systematically analyze every field within the client’s CRM system to determine which fields are statistical predictors of close (and which are not), allowing your data to tell the mathematical story. When sufficient data exists, individual reps’ and individual managers’ tendencies heavily influence our mathematical models.
We hate this term and we aren’t shy about saying so. There is absolutely nothing “secret” about our mathematical approach. If there was anything about our approach to sales data that our users didn’t understand, they wouldn’t trust the output. If they don’t trust the output, they won’t use the software. If they don’t use the software, they won’t renew their contracts. If they don’t renew their contracts, we don’t make money. And that would be a problem.
All of our users understand exactly what is driving the insights that our technology produces. As we learn about your unique data, you will learn about our models and how they will be adapted for you to be successful piLYTIX users.
Other tired clichés that we detest: “black box” and “proprietary algorithm.”
We feel that describing our offering in 21st Century buzzwords or shrouding our output in mystery insults the intelligence of our users and prospective users. Bad way to start a relationship, no?
There will never be any mystery about how our models work. When analytics companies are viewed as a magical “black box” solution, their output is always viewed suspiciously. By showing you exactly how our models work, you will feel more confident in the predictions and the prescriptions.
You can try and we wish you good luck. It’s probably bad form to answer a question with a question, but if you could just use your native CRM tools to solve these challenges, why are they still challenges?
If you are talking with us, there is something unique about your sales data. A one-size-fits all approach to predictive sales analytics, while relatively inexpensive, will underwhelm you. If your sales data isn’t all that special or unique, we’ll gladly take your money but chances are good that you will be overpaying us.
Also, lots of companies use “predictive analytics” to describe their approach. If a ninth grade algebra student could recreate the mathematical approach of other solutions, we might want to consider a different label.
See above. That’s a lot of work.
Lots of software companies make bold ROI promises. We are a math company at our core, though. We can’t claim ignorance of the difference between correlation and causation. Too bad! It would make for a nice marketing piece.
You were hired to grow your business and you will simultaneously pull lots of levers to achieve your desired outcomes. You will add sales staff, you will tweak your comp plans, and you will use other sales technologies. We will never have line of sight to many of those changes. Given that we can’t see – and certainly can’t measure – the impact of some of these other changes, it would be totally disingenuous of us to take full credit for all the growth that occurs while you are using our tools.
By the way, please be wary of any analytical solution that quantifies a guaranteed a return on your investment. It suddenly seems a bit shady, doesn't it?
Yes. If you feel that this is risky, we can help mitigate the perceived risk in ways that no other software provider will consider.
Definitely! If it’s your absolute final step in your evaluation process, yes. Until then, lets pump the brakes. Our clients are busy and they pay us a lot of money. We’ll use them for our commercial gain only when we are pretty sure that real commercial gain will result.
By the way, start tracking your own reference requests and see if you find any interesting patterns…
We won’t. But good luck. Traditional metrics of pipeline health are directionally fine. Until they’re not.
All of our users understand that there is some aggregate correlation between activity levels and win rates. Most of the people who hired us were hired by their companies to improve results. And so they want to go deeper than tired old industry metrics for pipeline health. They want to know which opportunities are mathematically most likely to close and which opportunities need management focus. They also know that if they coach and train to their reps’ individual strengths and weaknesses, they are much more likely to succeed.
Nope. It’s a really bad idea to do so. We know that this flies in the face of what those really expensive sales or operations consultants told you. Sorry! They gave you a bum steer. If you allow us just five minutes, we will be able to convince you how misguided their advice was.
Ask yourself has one of my sales managers ever said anything like:
“Ed’s committed deals always come in, but they never seem to be on time, so I will put half of them this month and half next month.”
“Diane is a sandbagger, but she always comes through in the end, so I will just put her at quota.”
“Bob is always overly optimistic, so I am going to knock his number down 25%.”
“Karen is typically spot-on with her forecast, but this deal has been in the pipeline for too long so I don’t believe it.”
“This prospect purchased at his previous company and loves us! This deal will definitely close!”
Unless your managers lack basic humanity, they, too are prone to biases. It’s not a good thing or a bad thing. But these biases (or blinders or tendencies – call them what you will) will prevent a manager from truly understanding potentially hidden strengths and weaknesses of reps and their deals.
piLYTIX was founded in 2013 to help global sales leaders address some of their most vexing challenges.
Successful sales pipeline management and sales talent management have historically been looked at as problems of an artistic nature. piLYTIX and its team of mathematicians see these issues as a series of interwoven—but eminently addressable—mathematical challenges.
Catering to companies and industries underserved by traditional analytical sales tools that tend to take a one-size-fits all approach, piLYTIX takes a consultative approach to onboarding and serving our clients.
piLYTIX works with your data as it exists and delivers to your sales organization deep, actionable insights into each opportunity in your pipeline and each sales rep on your team.
Jim Dries founded and serves as the CEO for piLYTIX. Jim's career began as an investment banker at BofA Merrill Lynch (formerly Banc of America Securities), working on more than $8 billion in equity, debt and M&A capital raising. After leaving Wall Street, Jim led sales, marketing and product management teams both at the Corporate Executive Board and Clubessential.
Jim has a B.A. from Yale University and an M.B.A. from the University of Chicago.
Kate Ivers serves as COO for piLYTIX overseeing all aspects of company operations. Kate's career has spanned law, operations and data security. She began her career as an attorney in the Intellectual Property group at Wildman, Harrold, Allen and Dixon. Recently, she served as the SVP of Operations and General Counsel for Clubessential where she oversaw product implementation, client services, human resources, legal and accounting.
Kate has a B.A. from Yale University and a J.D. from Loyola University (Magna Cum Laude).
Hendrik Kits van Heyningen serves as CTO for piLYTIX, overseeing all aspects of both the predictive modeling and the piLYTIX software product. Hendrik’s career began as an engineer at KVH Industries, Inc. (Nasdaq: KVHI), where he worked on R&D for inertial navigation systems, inventing and testing a novel approach to magnetometer calibration. Recently, he was employed at Analytics Operations Engineering, Inc., where he worked on scheduling and pricing optimization projects before ultimately taking leadership of the data science team that had developed the original piLYTIX models.
An accomplished musician, Hendrik has performed as a pianist at Carnegie Hall, and he served as Music Director for the Yale Davenport Pops Orchestra.
Hendrik has a B.S. in Mathematics and Physics from Yale University (Summa Cum Laude, Phi Beta Kappa).
You will be asked to argue a point during the interview that runs counter to conventional wisdom. Don’t panic. No need to qualify your position – just make an argument. We only hire people who are capable of creative and critical thought. This requires you to consider alternate and even contrarian positions before deciding on a course of action. If we are hiring you for any of the jobs posted on this page, we want creativity. Bring it. You won’t offend us. Our clients' success comes from the creativity and resourcefulness of our team. We hire talented, passionate people who should feel empowered to speak up and defend their positions. Debate is always encouraged. An occasional argument is fine if it gets us to a better solution for our clients. Lets make sure that you are comfortable with this during your interview. The secret interview question is: "What is the 9th digit after the decimal point in pi?"