Turbo charge your growth
Turbo charge your growth
Insight

Turbo charge your growth

Machine learning presents an exciting opportunity for mid-market companies to drive growth and outpace the competition.

Though the mad dash has already begun, it’s not too late to take advantage of machine learning to help grow your business and gain a competitive edge. Machine learning—or algorithms that analyze your company’s data and provide answers to critical questions—can be a game-charger for the middle market.

Answers to “How profitable will this customer be over time?” or “What campaigns best speak to our customers?” can provide valuable insight and help mid-market companies achieve necessary return on investment. Companies find even bigger returns by leveling up on the next round of improvements, securing a competitive edge in the market.

Machine learning gives mid-market companies a boost by:

  1. Finding more data insights than employees through repeated simulations
  2. Enabling companies to target specific segments in real-time
  3. Predicting more accurate impact assessments with large enough data sets. 

 

How does machine learning differ from traditional methods?

 

Traditional approaches


Intuition-driven, with  selective grounding in data

Identifies simple relationships  in small data

Straightline predictions  assume all else is equal

“One time” exercise that is  difficult to update


 

 

 

 

 

Machine learning approach


Fully data-driven

Identifies complex  multivariate relationships  in large data

Determines effect of  individual attributes  empirically and dynamically

Easy to update based on  feedback loops of new  observations


 

 

 

 

Benefits of machine learning approach


Identifies fact-based implications  a human would not think to look for

Scans a much larger universe of  potential signals to identify those  that are truly predictive

Creates more accurate  predictive models

Becomes “self-optimizing”  as additional information is  continuously made available 

 



Accelerate growth with machine learning in the front office

Machine learning improves decision-making and can quickly drive tangible results across marketing, sales and customer success for your mid-market business.
 

Marketing

Demand generation

  • Marketing ROI
  • Multi-touch attribution
  • Lead scoring/conversion

Digital performance

  • Digital presence
  • Social media sentiment

Sales

Sales effectiveness

  • Coverage model optimization
  • Discounting
  • Up and cross selling
  • Forecasting
  • Account rep attrition
  • Rep time allocation

Market performance

  • Customer, segment, and channel profitability
  • Multi-dimensional clustering
  • Competitive landscape mapping

Customer success

  • Retention drivers
  • Churn prediction
  • Renewal tactics
  • Preemptive maintenance
  • Knowledge and support pathways
  • Compliance

 

What can machine learning do for you?

Machine learning applies to a multitude of business functions, and a ripple effect can boost the bottom line for your business.  

Marketing

Marketing leaders in mid-market companies need help measuring the effectiveness of investments and making better decisions with new leads. Multi-touch attribution models can determine the impact of individual marketing investments on revenue and help mid-market companies switch tactics to see an increase in sales. Machine learning also can identify the highest potential opportunities with greater accuracy, allowing marketing to present the sales team with more productive leads.

Case study: A successful tech company sought to re-focus B2B marketing spend and increase the quality of generated leads without raising the marketing budget. KPMG developed a machine learning model that trained on marketing activity, while a targeted customer survey fine-tuned the parameters. KPMG discovered approximately 20 percent of the existing marketing budget focused on low-impact marketing activities and a shift of marketing spend to higher-impact activities showed a potential jump in gross revenue. 

Sales

With increased pressure to increase revenue, mid-market sales teams need help identifying which customers and leads to pursue and which offers and messages improve conversion rates. Machine learning finds data that provides a better understanding of consumer behavior, helping teams to increase retention, expand service and offer add-on purchases. It also can assist with revenue leakage reduction through focused discounting practices.   

Case study: An enterprise SaaS (Software-as-a-Service) company wanted to engage its customer base with new services to raise annual recurring revenue. Models created a 360-degree view of each customer, then scored and categorized upsell/cross-sell actions. Opportunities identified by the model resulted in a 2.5 times improvement on conversion rate, including a significant increase from high-performing customers. 

Customer success

Reactive processes and lack of visibility into customer behaviors plague mid-market businesses. Machine learning can help to identify likely deserters and downgraders, allowing customer service to engage these at-risk accounts and change the outcome. 

Case study: A SaaS company’s net revenue churn cost the company more than 10% of annual revenue. KPMG developed a machine learning model that could identify the signals that most accurately predicted churn and revealed customers at risk of churning up to 6 months earlier than previous methods. The team engaged these customers and improved annual recurring revenue by up to 4 percent. 

 

Five practices for achieving results with machine learning models 

When launching any new program, challenges arise, but we have five leading practices that increase impact and reduce time-to-results for mid-market companies.

1

Get the right people, right leadership and the right technology in place to answer the right questions. This includes close collaboration between marketing, sales and customer success teams, as well as data science teams, so machine learning models are created to address the most significant business needs. Secure quick wins to gain momentum.


2

Flexibility is key when starting, and learning by doing is the only way to master the power of machine learning. Teams should focus on a specific issue where a decision-making challenge exists, create a well-defined question to focus efforts, and go! Encourage the team not to think about what the data reveals but how it can be used to improve business.


3

Get everyone on the team onboard with machine learning tech, so they are comfortable and able to use it to the best of its abilities. Early stages should have teams finding quick, easy wins, allowing the front lines to improve their everyday performance and exceed goals.


4

Data, data, everywhere—use it now. There will always be more data to gather, but you can start taking action with the data you have collected. Successful teams seek data they can’t see, such as full pipeline history in user interfaces, which can be used to improve predictive models for revenue forecasting and marketing ROI.


5

Think big but start small with go-to-market approaches. Also, make sure to keep the field resources in mind, and adopt training and encouragement exercises as needed. Identify the champions on your teams (one on the business side and one on the data side) and unleash them upon your company with a tag-team effort to help the company outpace the competition.
 


How KPMG can help mid-market companies find their drive:

 


KPMG has experience working with mid-market companies—from $100 million to $3 billion in revenue—to establish data-driven growth programs.

We can help you tap previously untapped opportunities, focus prioritizes and build machine learning engines that will demonstrate tangible improvement potential in marketing, sales and/or customer success. With deep industry operating experience and functional knowledge, leading data science and machine learning capabilities and tools, and an approach that delivers results quickly and develops a culture of experimentation—KPMG can help mid-market companies succeed in today’s data-driven world.