Automation And Staffing Agencies

by: Veni Vidi Emi

We look at one of the most pertinent issues the staffing industry faces today.

We look at the impacts of automation and potential adoption timelines, all with a focus on the staffing industry.

The evolution of the labor market is of immense interest to almost every industry, but rapid change obviously makes projections difficult. We attempt to find the few areas with visibility.

This article is an in-depth look at automation's effect on the staffing space.

First, we introduce the general framework for analyzing the role of labor before moving onto describing how we will conduct our analysis on staffing agencies.

We begin by describing automation and what drives adoption. Then we look at projections that outline the scope and timeline of implementation. Those results are put in the financial context for equity holders and we conclude by outlining further areas of interest.


The labor market is a fun place to invest in for several reasons. The arbitrage between labor cost and revenue is usually quite small due to the low barriers to entry, little capital required (assuming proper working capital management), and general lack of scale properties. This leads to a vast array of undifferentiated competitors and a few winners. The few institutions that manage to position themselves for superior returns, either through complementary services, corporate structure and efficiency, or something else, usually have an identifiable advantage.

Researching whether this advantage will strengthen or wane provides an interesting (and potentially profitable) challenge. A concrete example would be Securitas (OTCPK:SCTBF), who complement their low-margin manned security vertical (similar business model to staffing) with a higher-margin business of selling security hardware. The job is made even more interesting by the accelerating pace of change. Labor has been a core concept in all economies since the dawn of civilization, but the paradigm shift since the industrial revolution has been the transition from manned labor to machinery.

With recent advancements in automation complexity and costs, the labor market is facing an intense change across the board. This leads to corporate strategy being absolutely central for any participant. I've worked on disruptees like Impellam (OTC:IGPPF), Kelly Services (KELYA), and Adecco (OTCPK:AHEXY), but also worked on certain bringers of disruption such as Blue Prism (OTCPK:BPRMF). There is however a third group. ManpowerGroup (MAN) and TrueBlue (TBI) being the primary examples.

Both have traditional staffing operations, but have attempted to diversify across new, but often closely related, business lines. In addition, there are staffing agencies who have ostensibly achieved this diversification, in one form or another, while retaining significant negative exposure to automation. An example would be Randstad (OTCPK:RANJF) and Recruit Holdings (OTCPK:RCRRF).

There are also executive search and management consulting firms, Korn/Ferry (KFY) comes to mind, which will largely remain unaffected. In the staffing world, the traditional players are like Chinese farmers and the technology players are like Genghis Khan. The outcome is all but assured, but the primary question is when the slaughter will ensue. The third group could be likened to refugees wise enough to flee beforehand. Stretching my analogy, a head-start and a sensible escape plan could work out wonderfully for stocks that most of the market has left for dead.

So let us take a deeper look at the boogeyman that the entire staffing sector seems to be afraid of.

Outlining The Impacted Areas

Before delving in, a framework must be put forth.

My personal understanding is that there are (roughly speaking) four buckets of spaces influenced by automation.

  1. Jobs that will be wiped out by automation in the intermediate term is a term for the types of jobs where recent innovation clears a pathway for final innovation that completely replaces the job. The final moat for many of these jobs is safety and cost. Many of these jobs are already seeing part-automation.
  2. Jobs that will see reduced supply in the intermediate term are commonly jobs where automation removes a repetitive (or entry-level) aspect of the job, but either requires or is enhanced by human interaction on the higher levels.
  3. Jobs that will be largely unaffected constitutes jobs where the supply and demand remains largely intact.
  4. Temporary jobs that will exist or face increasing demand as a result of automation is a fairly small bucket, but worth mentioning.

All these affect staffing companies across in three primary ways:

  1. Automation will act as a direct hit to staffing revenues from jobs no longer needing outsourcing, which leads to lower direct volumes from staffing.
  2. Automation will reduce ancillary revenue from recruitment process outsourcing, payroll control, job platforms, general HR, or labor consulting revenues.
  3. Automation will change the power-balance between suppliers and customers across all verticals.

This article will focus on the 1st and 2nd bucket and the 1st effect on staffing agencies.

The focus will be on the scope of automation in traditional staffing and the timeline for implementation, but with a material amount dedicated to the effect on professional staffing.

A quick definition of the two areas:

Traditional staffing is low-skilled temporary work and is often implemented on a seasonal, but contracted, basis. The work is lower margin for the staffing agencies, but has a higher margin profile.

Professional staffing is work that requires prolonged education, usually in IT, finance, or law. Professional staffing is often implemented on a project-basis and is therefore not as recurring, but has a higher margin profile.

Most staffing firms have a 80/20 split in their revenue constitution, with the vast majority coming from traditional staffing.

What Is Automation?

To quote from a previous article of mine:

"Automation is simply moving a process from being performed by labor to machinery using a set of rules or a repeating sequence of events.

Automation has been increasing for a long time with some of the primary drivers being the assembly line, the growth in robotics and computers, and more recently the improvement in sensor sensitivity and integration.

A classic example of manufacturing efficiency through automation would be a sensor that discerns (through color/size/etc.) whether a line of products being pumped out through manufacturing follow the guidelines.

The products that don't are automatically redirected. Previously a line of focused and observant laborers would have to manually assume this function, which not only increases costs, but also lowers consistency.

Below is a prime example of basic manufacturing automation."

Another type of automation is on an IT level, where easy data-input and retrieval has been replaced by efficient computer systems.

What Drives Adoption Of Automation?

The drivers of automation are two-fold: economics and permission.

Permission is based on stubborn unions protecting jobs and governments protecting their downside with both voters and social consequences. Permission is incredibly hard to forecast as it's at the core a human and political issue, driven by a complex range of incentives and personalities.

Economics drive adoption through reliability and cost. As costs per unit decreases through the adoption of automation, businesses will adopt the solutions. As solutions become more reliable, there are also fewer career risks associated with the transition.

What drives the economics of adoption is two-fold:

A commoditization of components and technological innovation.

As more firms begin to implement the existing technology and offer solutions, they will begin to compete on price and factories will have access to cheaper automation. Innovation will also drive the economic adoption through lower costs. Both of these are virtuous cycles as more adoption drives higher sales, driving both operating leverage and R&D.

As operating leverage kicks into gear, the cost per unit of automation components will decrease and through price competition lead to a lower overall cost for factories. Leading to broader adoption and further operating leverage.

R&D budgets will also scale, leading to even further technological innovation and cheaper and more efficient automation. As cheaper and better automation comes out, R&D budgets will grow even further.

Both are hindered by some physical constraints on chip size and efficiency at data centers, but the limits are far out.

What Is The Scope Of Automation?

The scope of automation isn't as binary as outlined in the buckets above. A study by the McKinsey Global Institute, linked here, describes the automation as a gradient from 0% to a 100% able to become fully automated at the current level of technology.

Source: Page 5.

According to the report: "Almost half the activities people are paid almost $16 trillion in wages to do in the global economy have the potential to be automated by adapting currently demonstrated technology"

An initial glance might lead readers to conclude that the impact won't be as large as predicted, but two points should remain at the forefront of your consciousness:

1. A job where 70-80% of the work is automated away practically removes the job or reduces the supply meaningfully.

2. Standard outsourced work is highly repetitive and fits the criteria for jobs highly susceptible to disruption. Furthermore there are fewer unions to protect the temporary workers in unionized countries.

But even without these considerations the graph is fairly telling. A vast amount of jobs will either disappear and be transformed into more advanced jobs.

According to McKinsey roughly 1.1 billion full-time 40-hour/week workers can be automated in the long run on a global scale.

Roughly 60 million workers in the United States alone are susceptible, which is equivalent to 1/6th of the population and roughly 25% of the current US workforce of ~240 million.

What Is The Timeline For Automation?

The key distinction in the timeline for automation is between technical evolution and adoption.

As we've seen on players like Blue Prism adoption and implementation takes time no matter the financial benefits.

The best estimates from the McKinsey Global report are outlined below, but I have a few comments.

Which is the more useful aspect? In my opinion "the automation potential timeline" is vastly superior information.

Automation potential is based on evolutions in technology, while adoption depends on competitive dynamics, "the permission factor" discussed previously, the evolution of global trade, and the economy.

The adoption projection forecasts that:

Our scenarios suggest that half of today’s work activities could be automated by 2055, but this could happen up to 20 years earlier or later depending on the various factors, in addition to other wider economic conditions.

But this is non-essential for staffing agencies. Staffing agencies deal with a specific set of jobs that have high automation potential and potentially accelerated automation timelines.

Let us outline the key areas of automation. TrueBlue probably has the best vertical-composition outline of any staffing company I've seen.


The bull thesis on each vertical is handily outlined by the presentation so let's focus on the potential negatives in the long run from automation.

Warehouses, Material Recovery Facilities, Low-Skill Sorting

A primary driver of low-skill labor supply is jobs related to grabbing and moving items as the variance in shapes throws off robot automation provided different shapes. This is why simple-form warehouses (such as Coca-Cola warehouses) have already seen full automation while less homogenous sorting jobs remain active across the United States.

Warehouses (in the retail & services segment above) have already been facing automation. Amazon (AMZN) has begun utilizing robots in their current warehouses, supervised by humans. The NY times contains some interesting, but essentially misleading information:

The paragraphs make it seem as if there were no real jobs lost, but consider two elements.

First of all, the workers had to undergo training courses. Spending money on educating workers is usually reserved for in-house employees, not a contingent workforce.

How many workers would have to work in a warehouse to equal the productivity of the new one? Amazon has essentially avoided building another warehouse with workers. I see nothing morally reprehensible in the move, but it is misleading to propose that massive productivity gains from machines will not lead to less supply of jobs in that sphere. The adoption timeline on warehouse automation still remains elusive, but with a majority of work in the largest warehouses already completed by robots adoption is largely to hit an inflection point quite soon.

Only three things are keeping a large amount of staffing work active.

1) The ROI on warehouse automation isn't excellent at the current level of innovation. According to this report a full automation has a 5-10 year payback.

2) The lack of ability to adjust resources deployed leads to an inability (utilization-wise) to manage seasonal peaks with economic efficiency.

3) Full automation is likely to lead to political uproar. It was very deliberate that Amazon decided to not lose any jobs at their facility renovation in 2017.

This leads to a few general conclusions. First of all, there is still ways to go before the economics make adoption necessary for competitors. Secondly, the adoption will not wipe out seasonal requirements in the short term.

The most important takeaway is that Amazon and other warehouse operators will attempt to adopt the technology by capturing the growth in activity, instead of replacing legacy workers. The legacy workers will continue to be used, but will likely be in-sourced.

The general conclusion is that warehouse automation could be adopted quite swiftly given the consolidated nature of the business along with a few improvements in ROI. Amazon is working on improving payback with their Amazon Robotics division.

In similar business operations such as material recovery facilities (where workers sort trash), the largest players have already begun to implement automated operations. I've covered the recent evolutions on that front in this article.

The projection of an adoption timeline is informed guessing, but given that:

a) Relatively unsophisticated players (waste operators) have begun to utilize automation in non-core areas.

b) The leading warehouse operators will begin to prefer in-sourced workers and use automated warehouses to capture what would have been increases in labor supply.

Any projection that doesn't estimate a material decline (+20-50%) in activity is likely too aggressive. For staffing agencies, one should expect +50% declines over the coming decade.

Construction Automation: Driven By Idiosyncratic Companies

Automation in constructing is a long-term worry for staffing companies.

Most of the work is non-active as of yet and driven by smaller individual firms. None of the work is comprehensive, but focuses on small bubbles of activity in construction work.

Given the slow rate of adoption on many of these issues, I will not cover the issue in-depth. For further personal research, I recommend looking into Fastbrick Robotics (OTCPK:FBRKF) and their Hadrian X machinery.

Logistics: Driven By Legislation And Large Companies

Logistics is another key space facing massive innovation and potentially quite urgent adoption.

Labor is an extremely large cost piece in trucking and other heavy-duty driving jobs (such as refuse collection). It will be extremely difficult to compete with firms that use automation while remaining manual.

At the same time, automation is being driven onwards by most major technology and auto companies. Ford (F) has recently announced a $1 billion investment in driving AI. Tesla (TSLA) is notorious for its development of automated software.

In the heavy-duty space, Volvo (OTCPK:VOLVY) is one of the primary contenders and have demonstrated their technology multiple times (1,2,3) and have the clear lead in refuse collection trucks (source here). Tesla is also a contender in this space.

There are five "levels of automation" as outlined by the Standard J3016 of the Society of Automotive Engineers.

A lot of projections have been made regarding the adoption of each step. A few sources for deeper dives into the technological evolutions can be found here and here.

The clear hindrance isn't technology, but rather regulation. There are millions (+1.8 million to be exact) workers in the US alone relying on trucking for sustenance. It will be extremely difficult to sign off on legislation that permits trucking automation, because unlike other industries, the status quo is that it is not permitted (due to safety risks).

And while automated driving is likely to be far safer, the news stories of even a few deaths is likely to rattle politicians.

The technology is largely irrelevant for an adoption projection. It simply hinges on legislation. If the innovations proceed at the current pace, the last remaining barrier in 7-8 years will be legislation.

Adoption will be extremely swift after a brief period for a few reasons.

First of all, automated trucks will be more efficient by far. Perfect braking, lower incidence rates, and optimal fuel use will all drive profitability. Seeing as depreciation and fuel are large cost-line items, the automated firms will be materially more efficient. Since trucking is largely a consolidated space with de minimis scale requirements, the price level will quickly drop to unsustainable levels for manual companies.

There are also cost-benefits entirely unattainable by average trucking companies with "good drivers." Platooning is when multiple trucks drive incredibly close to minimize wind draft. This is obviously incredibly dangerous for human drivers, but not an issue for computers. A study from MIT estimates up to 20% fuel cost savings.

Manufacturing: Idiosyncratic Across Verticals

Manufacturing is complex because there are a multitude of processes that differ drastically across companies. Many have been automated (which means less reduction from status quo) while other methods are still not cost-effective.

Since contingent workers often work repetitive tasks, I would simply use the aggressive adoption estimate from McKinsey, where most of the repetitive tasks have been automated away by ~2035-2040.

Conclusion And Other Areas Of Interest.

Looking across a traditional temporary staffing agency, it seems clear that almost every major vertical is facing major disruption. Construction the least, warehouses and manufacturing the most. Automation of logistics will be difficult to project as it hinges on legislation.

Overall it doesn't seem overly aggressive to assume LSD-to-MSD growth declines over the 2020-2030 decade for most temporary staffing. The slow transition will largely cut out the risks from negative operating leverage.

The point of this article is not that "staffing agencies will face -6.2353% growth given XYZ composition," but rather to illustrate that automation is coming (and to provide a general overview of how and when) so that investors will understand why even long-term flat growth projections in the staffing space is very optimistic.

The focus has been on low-skill temporary staffing, but as outlined at the start, other areas are facing disruption as well. Not only that, but the industry is transforming as a result.

Seeing as this article is already roughly 3000 words, I will not go on to cover all the other aspects, but I will recommend a few sources to read on.

A key aspect is to understand back-office disruption. I believe that my article on Blue Prism (where the automation intro was stolen from) provides a good overview.

Another key point is that automation will create new jobs and transform the industry structure. I would recommend a deep-dive into Randstad to understand how ancillary revenues might be impacted.

Disclosure: I/we have no positions in any stocks mentioned, but may initiate a long position in IGPPF, RCRRF, RANJF over the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

Editor's Note: This article discusses one or more securities that do not trade on a major U.S. exchange. Please be aware of the risks associated with these stocks.