Tuesday, May 16, 2017

Show Notes - Episode 7 - Managing Customer Service

Recorded: May 13, 2017
Host: Eric Veal
Guests: Ele Munjeli, Michael Cavitt, Andrew Sengul
Legend: Bold = key point

Chat 1 - Strategy

  • Intros
    • EM: devops engineer, JPL contractor, open source org "devopracy" and virtual democracies
    • MC: advisor and organizational consultant, helping leaders think
    • AS: SW dev, interactive narrative
  • EV: Strategy, management, measurement, is customer service dead and why?
  • AS: has it been killed by a new generation of services?  Venmo scams
  • EM: modern online businesses have much less cost. Reputation if bad can be resolved through re-branding.  Businesses more transportable.
  • MC: yellow page ads scams.  Bad service and scams is nothing new.
  • AS: scalability.  people don't scale and there are many humans potentially required.  
  • EM: expectations have changed.  People prefer automated service.  When is it appropriate to have a servant in a democracy?
  • EV: improved skills at delivering service through tech.  
  • EM: do people prefer or trust an ATM more than a teller?
  • EV: strategies can make customer service "dead" by demoting it and not having it be foremost.  Still examples of very high touch services.
  • EV: what was high-touch back in the day is not high-touch now (things change)
  • EM: two channels of people who prefer high touch and people who do not.
  • MC: "how much engagement you want" (depends on the customer).  DIY people vs. people looking to outsource something.  Pricing and services, product mix differs for each audience.
  • EV: markets and populations have people with both types of service needs (high and low).
  • AS: segmentation based on the significance to the buyer.  Segment based on if people see it as significant or not..
  • EV: don't just think about the human interaction part, consider the delivery and automation part other than just the human part.    Human touches diminishing over time?
  • EM: we may be replacing the "real" human interfaces with new human-like interfaces like voice.
  • EV: need to consider the customer service part from the beginning (common mistake).  
  • MC: need to build customer service experience vision into the corporate vision, mission values and long-term strategy.
  • EV: How do we include customer service during our strategic planning?  Looking at failure modes and the dark side?  Possible complaints, issues.
  • MC: need to consider risk and build a "ratchet" so you can go two steps forward but only one step back, for example (build in quality).  "The Luck Factor".  Write scenarios for the business.
  • EM: repair-driven systems design.  Program around areas where you know you'll have problems.  Build tools as you build products (comprehensive).
  • AS: Picture the failure from the eyes of the customer and design / build /release accordingly?
  • EM: Uber having issues now.  Transparent remediation.  
  • MC: Need things pre-emptively and to consider beforehand so we are more prepared?
  • AS: Humans encouraged to remove themselves from the sphere of customer interaction over time? 
  • EM: how do you build trust in automation? 
  • MC: customers really don't care, they just want their outcome.

Chat 2 - Managing customer service operations

  • EV: workforce, requests, and complaints
  • MC: Geico experience.  Sales and service tightly coupled.  Goal to resolve quickly with as few calls as possible.  People managed to reduce time on the phone.  Policies created conflicts.  You get what you measure in a lot of ways.
  • EM: Geico very pro-automation.  They have a non-human representative.  "Golden age of phone service".  More companies doing voice-to-text analysis.
  • MC: Hiring problem: no easy way to tell if a person is going to be good in customer service until you hear someone on the phone or see them in front of a customer.  
  • AS: Business and HR likes rote checklists
  • EM: wants more analysis on exceptional customer service.  What does it look like?  Predictive and anticipatory.  Are the best ones teachers?
  • EV: Skill required at many levels and a lot of risk.
  • AS: Human Competence book link.  People get stuck in 'cargo cults' who build a culture around checklists but have no agility, creativity or out-of-the-box thinking. What is the desired result?
  • EM: Humor plays an important role of moving from business-only to personal / deeper relationship.  
  • AS: The "Cute-ify-ing" of authority.
  • EM: Citizenship and extending government services.  Can't force adoption.  
  • EV: The weight of service delivery can be a lot when you are a consultant.  Working as a team way better, reduces the risk.
  • EM: Great delivery comes from empowerment.  Career "pathing" makes a difference.   What paths are available for people that start as customer service?
  • EV: People who start in the field can wind up in powerful positions.  But is this a common path?  Methods: 1) revolving door 2) isolated service org 3) one where people go to other departments.
  • AS: Institutions act primarily to further their own existence.
  • MC: Going from line / delivery into managerial roles is not necessarily good.  Army did specialist ranks.  Managerial routes don't always make sense.
  • EM: Tech support should get into UX at some point.  Need a closed loop.  Tech support part of the design process, instrumental in designing the requirements, for example.  50% of problems are usability problems.  

Chat 3 - Measuring customer service

  • AS: Cargo cults further explained.  Tribes would keep hoping for the planes would keep coming.  Doing non-value adding things and hoping.  Opportunity to find low performers and bring them up (low hanging fruit).  Need to set priorities and focus.
  • EM: What and how are you measuring success and failure.  What is a meaningful failure?  Some people rewarding the wrong things and behaviors (like people fixing bad issues)...as it encourages more bad behavior in the future.
  • MC: Celebrate the architects or the recovery team?  Information overload can be bad.  Not good to just get all the info.  Know the questions.
  • EM: Designing good metrics is a trick.  Need to avoid bias and ask good questions (need good science).  There's a wrong way to ask questions and gather info.  Witch hunts, for examples, not a great way to gather info or do science.
  • AS: Absurd metric examples.  "How slurred is their voice?"
  • EV: Instrumentation and telemetry a big part of this (more is good) but still need filters and need to roll it up, prioritize, control, etc.  But more data generally a good thing.
  • MC: Text-to-speech analysis is important and good feedback source.  Changes many things.
  • EM: The relationship and rapport allows us to get better data.  Have users participate in the process.
  • EV: Cost of getting feedback reducing through better instrumented products.  Doesn't require customers to actively give feedback but do through use instead.  Build the feedback system right into the product like a vacuum could detect it being kicked or sworn at.
  • EM: Want a closed feedback loop where the feedback you gave was finally told to you that your comment mattered, when and how.  
  • AS: Investment and significance.  
  • EV: Investment happens on both sides: the customer and the provider.  Both need to be invested and quid pro quo may be needed.
  • AS: Speculation and investment.  A financial question.
  • EV: Human capital is next.