The Transformative Power of Algorithmic Resource Forecasting

How Shinto streamlined Wunderman Thompson Health’s resource allocation and scoping workflow, saving over $1.3M in three years


For a downloadable version of this case study, click here.


Shinto is a Software as a Service designed to provide project managers and financial decision-makers with low-effort yet highly accurate resource forecasting.

In 2018, Wunderman Thompson Health (WTH) was looking for a solution that would unburden its higher-level project managers and financial leads, who spent inordinate amounts of time and effort on what was ultimately ineffectual resource and financial forecasting exercises.

Once Shinto was chosen to solve this resourcing problem, WTH began offloading the responsibility of resource forecasting onto Shinto’s algorithms, which analyze both a company’s timesheets and general ledger. Shinto’s probability waves map the future activity of all team members over the course of all current and planned projects, providing an accurate view of resource availability which guides hiring decisions across all departments at WTH.

To date, Shinto’s efficiencies have saved WTH over $1.3M.

The Challenge

For years WTH has maintained a robust Project Management division with well-documented processes and thorough financial records.

However, until the introduction of Shinto in 2018, the mechanisms for creating budgets, scopes and schedules weren’t automated or integrated, and much of the effort put into meticulous financial record keeping couldn’t be plugged into anything that “learned” from it. This meant that every year thousands of man-hours were being expended for repetitive tasks that were better suited for intelligent automation.

Further, WTH’s resourcing and resource forecasting on the named-resource level was not only extremely labor-intensive, but also extended only two weeks into the future. This was in stark contrast to their hiring processes, which could often take several months, which is not uncommon for most advertising agencies.

Finally, scopes and budgets were essentially drawn up by hand, requiring hours or days to complete. RFPs and pitches were incredibly arduous, demanding high levels of highly-skilled resources in a short period of time, and were very prone to human error.

In summary, WTH was committing extensive human resources to budgeting, HR / financial forecasting, and to fine-tuning and adjusting scopes and pitches over and over, essentially by hand.

This was a prime opportunity for the integration of an intelligent tool to provide powerful, efficient forecasting.

By incorporating Shinto, WTH saved approximately $250,000 per year in unnecessary full-time hires in the Washington office, then more than twice that when they onboarded Shinto to their New York office in year three.

This total savings of >1.3M USD over the course of three years, and in only two of WTH’s branches, dwarfed the costs of licensing the software, and forever changed the way WTH’s project management and financial teams approached resource forecasting and project scoping.

The Solution


Before WTH first began collaborating with Shinto, they were generating each individual budget from scratch, even for repeat projects for the same client. The manual process involved Excel spreadsheets and NetSuite OpenAir.

In order to streamline the budgeting process, Shinto helped the WTH Project Management team scrub the financial data readily available from their existing timesheet system and general ledger. This meant sorting and identifying their ten most-common project-types, then averaging the amount of hours burned by the most common job titles that worked on each project-type.

A secondary, deeper dive was then conducted to identify variances in the hours burned on each project-type between different accounts. Shinto’s unique ability to automatically produce reports that show how much time it takes for each role to work on each type of project proved invaluable.

Prior to Shinto, this enormously useful timesheet and ledger data had been left unutilized, despite the effort put into meticulously reporting it. Shinto was able to mine it in order to provide for better-informed budgeting and scoping. This produced a catalog of budgets for all typical project-types at WTH, plus variants of each typical budget for each of the various accounts at the company. This data was catalogued for input into Shinto in Phase 5 – Company Setup.


Financial forecasting can be exceedingly challenging at advertising agencies, just as it is for any other business whose work volumes are highly variable and unpredictable. More often than not, most analysis at firms like these is conducted in retrospect—reviewing last month’s or last quarter’s results, as opposed to forecasting and preparing for the future. In the absence of data-based forecasting, many firms have no choice but to rely on guesswork or the anecdotal experience and impressions of individual managers.

WTH was no different in this regard. Upon identifying this pain, the Shinto team returned to the timesheet data used to develop the budgets in Phase 1.

Now, instead of analyzing total hours per job title and project- type, Shinto set out to determine when hours were burned, throughout the lifetime of a project, according to job title and project-type.

Projects could now be created in seconds, often
by junior staff, because the company’s knowledge now lived in the software.

For example, if a User-Experience Designer typically billed 32hrs on a certain type of project on a given account, when did they bill those hours? Just as before, all the data required for this analysis was already in WTH’s pre-existing timesheet software and general ledger, but it had only been utilized to analyze the past, and not to forecast the future.

Shinto was able to identify the phases in a project’s duration where each job title generally billed most of their budget allocations on the typical project-types that had been identified in Phase 1. In the example of the User-Experience Designer, it was found that when working on a website project, they generally billed 65% of their hours in the first 3/10 of a project, completing wireframes. They would then spend another 25% of their allocation in the 4/10 of the project completing functional specs. Finally, they would spend the remaining 10% of their hours in the last tenth of the project participating in User Acceptance Testing and Quality Assurance (QA). This type of analysis was refined further when Shinto identified different trends between accounts/clients for certain resources on project-types.

This data was also catalogued for input into Shinto in Phase 5 – Company Setup.


This phase is where Shinto starts paying for itself many times over on a monthly basis.

Once Shinto had access to historical budget allocations for specific project-types, it began to learn where in time those hours could be expended on each type of project. This armed Shinto with the ability to forecast hourly fees as they fluctuated according to the projected work schedules that ebbed and flowed over the length of each project.

At this point, projects could now be created in seconds, often by junior staff, because the company’s knowledge now lived in the software. Once the projects were created, Shinto automatically shaped work-hours across time for all the titles on each project, and those hours were applied to bill rates, resulting in weekly, monthly, quarterly and annual forecasts for every project and account.

Those same hours shaped across time were simultaneously compared by Shinto with the capacity for each department, title and named resource vs. capacity in the time period the user selected, including each individual’s personal leave—this was an impossible achievement before Shinto.

It should be noted that in the accompanying data (“Results”) that there is no comparison between the time required for financial forecasting by hand and that required by Shinto. The reason for this is simple: resource and financial forecasting across time is an automated corollary of creating each project within Shinto. Shinto’s algorithms draw upon the project parameters automatically, meaning that the only time required for forecasting was during the initial project set up, which, as stated above, can be done by junior staff members.

As described by Scottie Lee, WT Ops Lead, PMP,

” Shinto made us analyze ourselves in ways we hadn’t before; in order to setup Shinto we had to truly look at what it really costs the company to make things — who takes how many hours and when it is that they use those hours chronologically.

“ Now Shinto knows those answers and can forecast based on that information, so my team doesn’t have to. I can literally put a rookie PM on a pitch team and rest assured that they are pricing projects out based on real data. “

Shinto’s resource and financial forecasting can be reviewed, adjusted and exported as often as necessary with minimal effort.


Most advertising agencies have very limited options when selecting scoping software, and end up resorting to drafting each scope manually. Shinto’s ability to dynamically generate a project SOW drastically reframed WTH’s entire scoping workflow.

The Shinto team began by reviewing past WTH projects and compiling SOWs for the typical project-types identified in Phase 1. Next, WTH identified the SOWs that WTH felt best represented each project-type. The best language was lifted from each SOW to create high-quality SOWs for each project- type.

Resource and financial forecasting
across time is an automated corollary of creating each project within Shinto.
Shinto’s algorithms draw upon the project parameters automatically, meaning that the only time required for forecasting was during the initial project set up.

(The undertaking of this exercise had ancillary benefits as well, as it led WTH to write and formalize language that better protected them from the poorly defined scope phrasing used prior.)

The second step in the scope-generation phase was based around the experience and knowledge of their very qualified Project Managers (PMs). The PMs were asked to determine the level of effort (LOE) for incremental changes in a scope, broken down to the individual title level. For example, how many hours of additional Developer hours would generally be required if an additional round of QA testing was required on a banner project? These numbers were catalogued for inclusion with their corresponding scopes.


Before approaching Shinto, WTH had previously onboarded an enterprise-level Project Management software system. The integration process took months, draining resources from the Project Management team, and required heavy involvement from the Finance team. Expensive consultants were contracted to usher the company through the overall process, as the vendor of the Project Management tool did not offer integration services.

In contrast, integrating Shinto took no more than a couple of weeks, the bulk of which was analyzing prior budgets, with the onboarding consulting services being included as part of the subscription package.

As Kimberly Cassidy, SVP, Project Management & Agency Operations, Wunderman Thompson, PMP describes it,

” One of the nightmares about bringing on a new enterprise-wide system is adoption—people usually don’t want to try something new, even if it’s better than what it replaces, but Shinto’s integration was so effortless that nearly all of the system onboarding could be done by people who weren’t even PMs. Basically, anyone with a PC can set a company up from scratch.

The Company Setup Phase took two days of simple data input for things like employee names, titles and bill rates. The most time-intensive task during this phase involved aiding the PMs to scrub their existing data.

By design, Shinto’s Admin Interface requires the absolute fewest steps to get up and running. Those steps form a logical progression using the user-friendly Admin Interface, resulting in a system setup process so simple that most of the integration work was actually done by a junior staffer at the front desk with no formal project management or finance experience.

It was during this setup stage that the data compiled in the previous phases was simply input into Shinto: the typical project-types, the corresponding scopes and hours per resource required for changes to the variables in those scopes, the phases in each project where a given resource typically bills their time, client names, employee names, bill rates, etc… This one-time task by a junior staffer replaces the recurring, uncountable hours of high-level data massaging that was previously required for each project.


Often the training required for using an enterprise-wide project and/or financial management system is weeks long, or longer. And that brings its own cost, both in direct costs as well as productivity loss.

Integrating Shinto took no more than a couple of weeks,
with the onboarding consulting services being included as part of the subscription package.

Basically, anyone with a PC can set a company up from scratch.

But because Shinto is built to be a fewer-click forecaster, which gleans data from Project Management and Financial, the entire Project Management team at WTH was trained to use Shinto in one day. PMs of every level were trained in creating a project, which on typical projects takes no more than a few clicks, instead of the hundreds and thousands of clicks required previously.

The second block of the training centered on analyzing the resource management and corresponding financial forecast that results from the project data.

The third block of the training, encompassing automated scope of work creation and adjusting, took an hour.

In the fourth and final block of this 1-day training, senior team members were granted authority to override Shinto’s automated forecast, in order to personalize the budgeting for projects deemed outside the norm. The team was then instructed on ways to train Shinto to adjust the baseline numbers for certain project-types on their particular accounts, so the next time they opened a given project-type for a given client, Shinto would have an updated budget based on more recent historicals.

In order to reinforce this training, Shinto’s interface has built- in FAQs which include many short explainer videos, providing in-depth explanations of each element.

The training sessions were attended by dozens of PMs, the CFOs, Controllers and Accounting Directors.


Once Shinto had access to the data from hundreds of projects, it was able to provide the WTH team with novel views into financial, time, and resource forecasting that they had never before been able to access.

This resulted in the creation of new processes and procedures for how and when the data would be used, as they could now make hiring decisions supported by data-driven forecasts.

For example, the Account team often reaches peak demand in mid-summer. This might normally prompt full-time hires for Account Directors. But with Shinto, WTH was able to detect a dip in demand for this post in Q3 and Q4. By opting to hire freelancers to cover the mid-summer surge in place of full-time employees, WTH was able to save hundreds of thousands of dollars in salary, a savings far exceeding the cost of Shinto’s software subscription.

These forecasts — presented in a format specifically designed for advertising agencies and the creative service industry — give resource managers and department heads the ability to not just analyze their future needs, but also provides the finance team and CFO the ability to look at how billable a team, title, or employee is as compared to salary, providing even more actionable data points.

The PM teams were able to budget, scope and resource a typical project in a minute or so with just a handful of clicks compared to the hours and days (and thousands of clicks) that their prior methods required.

Further, the ability to budget and scope dozens of projects in an hour, which are viewable in a client-friendly Gantt chart produced by Shinto, enabled them to have fewer and less- senior Project Managers work on new business pitches, and at a dramatically faster speed. This, in turn, allowed the business development / pitch team to work more in-depth on concepting and planning with newly-found time, because it now took less time to scope out and price the work.

Retrospective reviews of Shinto’s first years of resource forecasting reveals an accuracy threshold of <10%. With each passing year of aggregating company-wide data, Shinto’s algorithms become more and more accurate and effective.


By integrating Shinto over the course of a few weeks, WTH was able to save over 1 million $USD over the course of three years, a sum that is orders of magnitude greater than the price of the Shinto software license.

The accurate guidance provided by Shinto continues to allow WTH to base its hiring decisions on hard data in order to avoid unnecessary full-time hires, or, alternatively, to prevent the long-term overdependence on freelancers.

Additionally, the simplicity of the tool frees up high-level positions from tedious and ineffectual forecasting exercises, providing instead valuable, actionable guidance on future hiring strategies, with a minimum of data input.

Finally, Shinto has streamlined WTH’s scoping workflow, making it more flexible to modifications on the fly, and allowing low-effort generation of high-quality SOWs.

A downloadable version of this case study is available here: