Every Company Has an Undiagnosed ‘Last Mile’ Problem

We’re drowning in business operations apps. So why isn’t anything easier?

Sagi Eliyahu
8 min readJan 13, 2020


Photo: Oliver Le Moal/iStock/Getty Images Plus

IImagine the following. It’s 4 p.m. on a Friday, and the client project your account manager has been working on all afternoon is almost finished. She’s already completed 80% of the work; all she needs to do is confirm deployment with engineering, get sign-off from legal on the contract, and confirm final logistics with you, the VP of Customer Success, to whom she reports. Then she’ll be able to close her laptop, leave the office, and return home to her family for the weekend.

Unfortunately, that’s just not how things work today. The people she needs to complete these last steps work in separate, siloed platforms. Her director of legal ops only uses email; the engineer she’s been working with operates in Jira; and your team works in Salesforce. Because these packaged applications can’t communicate with or adapt to each other — they, in fact, weren’t designed to be adaptive at all — the onus of orchestrating this final 20% of collaboration falls squarely on her.

Why is this a problem? Well, for as technically powerful as these applications are, trying to use them collaboratively requires a beguiling amount of human effort. Across a veritable honeycomb of open browser tabs and blinking application windows, employees are forced to chase follow-ups, pester colleagues for signatures, update a variety of systems, and manually input pages of data into CRMs. All told, it’s remarkably inefficient, not to mention draining. By the time your manager of customer success has finally updated all of these systems and coerced all the various stakeholders to complete their necessary functions, it’s 7 p.m. Her partner — at home with the kids — is upset. Because you wanted to ship the completed project to the client by EOD, you’re also annoyed; the company has lost money. And as she closes her laptop for the night, she, too, feels resentment, both for the operational inadequacy of the applications you buy, and for the expectation that she has to be the one to compensate for their limitations.

This scenario exemplifies the “Last Mile” problem in business operations

And it can be boiled down to the inability of the systems and applications we rely on today to complete processes or projects end-to-end. Just as mass transit systems can only ever transport a commuter 80% of the way to their office — dropping them off at a train or subway station roughly one mile from their office, for example — the enterprise cloud applications we all use today can only ever complete 80% of a given job. They just aren’t capable of meeting each individual’s differing door-to-door needs.

This inefficiency is currently wreaking havoc on companies and employees alike. As Mary Lacity and Leslie Willcocks wrote in the Harvard Business Review, because of the Last Mile problem, “Workers must now spend substantial time dealing with… quirks and shortcomings that are just as endemic to systems as their strengths.” As a result, employees spend only 2.8 hours a day completing high-value work. Being forced to dedicate such an immense amount of time to mindless tasks lengthens workdays, and equates in the minds of employees to poor leadership, which has been found to be a leading cause of burnout. It goes without saying, meanwhile, that such inefficiency is horrible for bottom lines.

For these reasons (among others), the Last Mile problem very well may be the most pressing issue company executives face today. Unabated, it threatens to inhibit the innovative capacity of every company operating today, and further erode employees’ already-diminishing quality of life.

Something needs to change.

A monster of our own making

The “Last Mile problem” as a concept first gained traction in supply chain management. Specifically, it was used to reference how the last link of any given supply chain — transporting a product from the freight station to a retailer, for example — always proved the least efficient, comprising up to 28% of the total delivery cost. This adds up to more than a trillion-dollar problem globally. It’s become more common to use the term today in the context of transportation — namely, in reference to the aforementioned inability of mass transit systems to transport commuters 100% of the way to work.

The Last Mile problem in business operations, another trillion-dollar problem, is a similar but different beast. While the inability of public transit to carry us all the way to our desks is inherent — a matter of geography and inalterable urban planning — the Last Mile problem in business ops is self-inflicted, a monster of our own making.

It was born at the same time cloud software — that game-changing innovation initiated by Salesforce in the early 2000s — unshackled companies from the expensive limitations of on-premise software and the cost of developing homegrown solutions. And while history typically refers to this momentous event as a fundamentally positive thing, it had at least one negative side effect. Namely, it birthed the pernicious mindset that nearly all companies maintain today and that has in turn allowed for the Last Mile problem to explode: to solve a problem, one should merely buy an app. Need help managing customer relations? There’s an app for that. Need help chatting, assigning tasks, or sending invoices? Yep, there are apps for those things, too!

In the beginning, these apps we aligned ourselves with were genuinely beneficent. But we didn’t understand what we were committing ourselves to, the kind of Frankenstein we were enabling. We were ecstatic when an application increased efficiency in a given vertical 80%, and we were even more pleased by the speed with which SaaS products got to work. “What an improvement!” we thought.

We proceeded to turn these applications into the foundational pillars of our businesses.

The trouble was, no one realized the operating cost of saddling employees with the responsibility of compensating for that remaining 20% — of asking employees, over and over again, to mold their behavior around the needs and limitations of each new process. To compensate for the “Last Mile” with one new app is a mere annoyance, but asking employees to do so across a bustling and increasingly essential ecosystem of platforms and apps means encumbering their productivity with an ever-expanding amount of manual work.

Nevertheless, we proceeded to turn these applications into the foundational pillars of our businesses.

Which brings us to the world we live in today. On our laptops and in our pockets, we all enjoy access to packaged pieces of software that are cheaper and more powerful than ever before. But our employees spend more time compensating for their limitations than they do benefit from their functionality. And because our preferred means of problem-solving remains outsourcing problems to enterprise apps, the number of apps we use continues to proliferate — meaning the number of apps we force ourselves to adapt to and change-manage (with only marginal improvement) also continue to proliferate — thus perpetuating the Last Mile problem into eternity.

In other words, we’re locked in a sick, self-reinforcing cycle that hurts both our businesses and the people who power them. And unless we find a different way of designing mission-critical processes — preferably one that puts people first, and that empowers workers to collaborate symbiotically with technology — it will only get worse. Employees will lose more of their time to mundane, compensatory toiling, and companies will continue to lose the ability to innovate.

The solution? Think differently.

As I hinted at above, the solution to the Last Mile problem in business operations centers around this idea of putting people first — of designing processes and systems that serve and empower employees. As such, we need to seriously rethink how we design and use apps. The apps we use to run our businesses should not only complete one siloed function — rather, they should be capable of adapting to the user and their needs. The apps we use should be personalizable, capable of cognition, even, such that we can use them proactively and collaboratively to solve the Last Mile problem in each of its unique contexts.

Now, there are automation solutions available today that focus on integrating systems or automating repetitive system-based tasks. But that’s not what I’m talking about. Those solutions only address a portion of the problem. Even robotic process automation (RPA), which once promised a veritable revolution in automation, only addresses limitations in legacy software — which is awfully short-sighted.

What the world needs instead is a solution that’s truly end-to-end in scope — that works with you to automate mundane, time-consuming tasks and that’s capable of managing processes and workflows across systems and people. We need a tool that dismantles the barriers to innovation and productivity, a tool people designed to eliminate work, not create more of it.

What we need, in other words, is software that’s adaptive — that operates like a powerful, dynamic virtual assistant

Again, imagine your same account manager, preparing to complete the last 20% of her project. Imagine if, instead of her having to traverse that last mile on her own, she could employ a powerful, adaptive personal assistant to do that work by interacting with the systems and people in her organization for her, behind the scenes — and that could even anticipate her operational needs so as to be proactively helpful, saving even more time.

In such a case, instead of having to complete that mindless work on her own, she could spend time doing work befitting her intelligence and skill.

Now, the idea of such a tool might strike some as science fiction, but such game-changing technology is in fact already available. But for it to make a tangible difference in the way we live our lives — similar to the difference the cloud revolution made, for example — the availability of powerful, adaptive software isn’t enough. All of us have to change the way we think about building and implementing solutions. Instead of shopping for apps that each only solves one specific problem and that employees have to accommodate personally, we should demand more holistic, customizable solutions that work with our people.

This shift amounts to nothing less than a revolution in the way we build and operate businesses. And a revolution might feel like a tall order. But it’s necessary, and more importantly, it’s within our grasp.

To start, try the following. Next time you have a business challenge, and you feel that familiar urge to search for an app to solve it, pause and ask yourself: Is this app going to create more or less work for my people? Is it going to perpetuate the “Last Mile” problem, or solve it?

With any luck, just as the idea of purchasing on-premise software feels so ridiculous today, tomorrow, the idea of solving problems by purchasing more packaged apps will strike us as similarly preposterous.



Sagi Eliyahu
Writer for

Founder & CEO of @Tonkean; An entrepreneur, innovator and tech guy