Chasing Perfection: How ShipBob Drastically Improved its Order Accuracy Rate

Fulfillment is like heart surgery: People appreciate successful surgeries, but if a surgery fails, no number of prior successes really matters. Perfection is expected.

At ShipBob, we take our clients’ shipments very seriously. We know a ton of work goes into making a product a reality, from sourcing materials, manufacturing, and shipping the goods – often between countries – to arriving at our doors and fulfilling orders. ShipBob’s hands are the last to handle these goods before they are sealed in a box and delivered to the end customer. It’s extremely important that we do our job perfectly!

We stow, sort, pick, pack and ship tens of thousands of shipments every single day. Given this volume, mistakes are bound to happen. Even the best people are not perfect. Nonetheless, perfection is expected of us.

In full transparency, it’s easier writing this in hindsight, and our starting point seems very rough. We had to make some extremely hard decisions along the way.

At one point, our error rate was hovering around 0.5%. Considering the fulfillment industry average is 2.0% and the best practices/industry benchmark is 0.5% (F. Curtis Barry & Co.), we could have stayed comfortable there. However, this meant that 5 out of every 1,000 shipments had some sort of error. What are errors? An error is any defect to the expectations for the shipment. Some example of the defects that ShipBob takes into account relate to weight, processing time, custom packaging instructions, SKU quantity, packaging dimensions, among others.

Ever received a box with way too much empty space?  We would consider that an error.  At a high level, if a shipment does not meet the expectation in any way, it is an error. Fulfillment is like heart surgery, and perfection is the goal. Read on to learn how we set out to achieve a drastic improvement in shipping accuracy.

The journey to perfection and 5 steps to getting closer:


Step 1: Setting a Clear Goal

We started with an ambitious but attainable goal. We set out to achieve a 5-sigma error rate, which is equal to 2.3 errors per every 10,000 orders (or a 0.023% error rate).  To spice it up, we decided to accomplish this goal over the course of three months and to do so without any significant technological advancements.  In other words, we wanted to get 20 times better in roughly 17 weeks. These goals were clearly communicated to everyone in the company and continuously reinforced.

Step 2: Building a Scorecard and Accountability

If we were going to make something matter to anyone, it had to be clear and trackable for everyone. Thus, we made a scorecard for each team member that tracked the total orders and actions they worked on and their error rate. We reviewed this daily. To emphasize how important this initiative is, we made these scorecards available to the everyone in the company. The goal was to make everyone’s work more transparent, and it allowed us to set very clear expectation at the individual-level.

We set the individual baseline expectation of a 0.4% error rate as a starting point and refined it over time. If a person dropped below it, they were put in the red zone and placed on a performance improvement plan.  Simply by making this change, the organization and individuals started greatly improving.

Takeaway: Most people really care about their work performance, and, as a leader, you should shine a light on it. Making people’s work and results transparent shows you care. Most people will step up to the plate and take greater pride and responsibility in their work when given the opportunity.

Step 3: Incentivizing Improvement and Making it Matter More

To attain a goal as group, each team member’s motives must be aligned.  To keep it simple, we created incentives around monetary (time) compensation and recognition for actions that were in line with the goal. We tried a bunch of different approaches, but here is what stuck:

  1. Incentive to prevent errors: Functions A, B, and C are consequential. If the operator in function B caught function A’s mistake, they got a monetary reward.
  2. Incentive for individual perfection: If a person completed error-free work for four weeks in a row, they received a bonus. If they did it for eight straight weeks, they received double the bonus.
  3. Team perfection incentive: Each function/department receives a perk for going a week error-free (e.g., comped lunch or money to buy something for the break room).
  4. Public praise for the success: People and teams were praised like all-stars during group meetings, the time at which we handed out any earned rewards.

This combination gave tangible motivation to catch errors, excel individually, and excel as a team. With these efforts, we reached an error rate of 0.15% and made significant improvements.

The team got three times better in under three weeks. This was pretty spectacular, but only got tougher from there.

Step 4: Instilling a Culture of Perfection

To be able to expect perfection out of our staff, we had to make it a living and breathing value within the organization. This is difficult to make tangible and actualize, but it really began on day one.

How do you make people value perfection in their work and their environment? I would be lying to say I have the full answer, but we made progress in right direction.

As soon as our goal of 5-sigma was made clear, it became the seed of the culture of perfection and we planted it in everything we did, from the top-down.  Every meeting, every one-on-one, and every all-hands started with our goal in mind.

To those who thought they were already doing their best, it seemed a bit unattainable to get significantly better. Thus, we set up milestones on which to amplify the cultural values of perfection along the way.

The first milestone was for the team to have a perfect week. It only took two weeks for the first team to achieve it, and that team was praised like the all-stars they are. The next milestone achieved was for a staff member to have a perfect four weeks.  With each milestone, people got more motivated and it generated more buy-in. They started to really look out for each other, and it made the cultural value tangible and real.  People can buy into a company culture if they can feel it and see it as a reality.

Getting below 0.1% consistently as group was another milestone for the team to reach. Within 10-11 weeks, we hit that goal. I can comfortably say that at that point the team was living the culture of perfection.

Step 5: Making the ‘A’ team

This was the hardest step. Along this journey, it became clear that certain people were stepping up to the plate and others were not.  We were raising the bar for a low error rate and the company as a whole. Some people were simply not up for the challenge.

Expectations were clearly set and a straightforward process was created to achieve them. If people are not achieving, you ultimately have to let them go. Did I mention this was tough? We were building the “A” team, and it was obvious that some people were not up to it. However, some people were good and some days they were great, but they were not great all the time. It was extremely tough but absolutely necessary to let these people go as well. You have to be clear and consistent. There is no gray area here. People either do great work or they don’t. There isn’t an ‘almost.’

I would like to think that we did this in a fair manner which made it easier to accept. Metrics and goals were clearly defined. If people got in the red zone, they were put on a two-week performance improvement plan for the manager and team to work with them. Each person had control over their results. Nonetheless, we had to let some people go.

Results

Overall, this process achieved great results. At the end of the three-month period, we achieved an error rate of 0.052% (or 5.2 errors in every 10,000 orders). Two weeks later, we hit our goal of 5-sigma and a 0.023% error rate was achieved. This was incredible! After that, we didn’t just cross the mark, we stormed passed it and achieved an error rate of 0.011% (1.1 errors in every 10,000 orders). In a span of only 19 weeks, the team had gotten roughly 50 times better without any major technological improvement! It was time to have a barbecue and break some bread on the achievement.

But then again, ecommerce fulfillment is like heart surgery and one in 10,000 failures can still be improved. At ShipBob, we achieved a significant milestone in a relatively quick timeframe. We inched closer to perfection and continue to make gains in the chase. We will never stop striving to provide the best possible experience for our customers, and we look forward to continuing on this journey.

Learn more

See if ShipBob is a good fit for you by talking to a fulfillment expert. Request a pricing quote below.