Goldman Sachs’ analysts produced a research report in April 2019 that examined why stock investors undervalued startup businesses that had been built inside larger firms. 

It was six months after David Solomon had become the bank’s CEO. He was growing increasingly frustrated by the fact that Goldman’s stock price didn’t seem to reflect the promise of a digital consumer bank that Goldman was building. 

The analysts wrote under the banner of an in-house think tank, at times used to advance the goals of the bank’s executive suite. 

This time, however, the analysts handed Solomon a rare rebuke. Led by the then head of research, Steve Strongin, the analysts laid out a convincing argument for why investors were right to be skeptical. There is a real chance, they wrote, that management teams will fall in love with a pet project and throw good money after bad. Independent startups, on the other hand, have a natural cap on their spending that comes from having raised a finite amount of money. 

The analysts gave it a name: “deep-pocket risk.” 

There’s now growing evidence that the shadow analysis may have been correct. More than five years after launching its first product, Goldman’s digital bank, Marcus, is burning cash with no signs of turning a profit. Executives expect it to lose another $1.2 billion this year, Bloomberg recently reported, after Goldman publicly predicted that its consumer business would be in the black. 

The ballooning expenses have attracted the scrutiny of Solomon’s No. 2, John Waldron, who now holds regular meetings with Marcus leaders in an effort to get costs down, according to a former employee who has been briefed on the meetings. The elevated costs have frustrated Waldron, the person said.

If it was just a matter of expenses, Marcus’ problems might be manageable, according to roughly a dozen former executives, junior employees, consultants, and others familiar with Marcus who were interviewed for this story. 

But the dysfunction extends far deeper. The business has failed to knit various products into a coherent strategy, with some, like a robo-advisor, struggling to gain traction and others missing deadlines for public launches. A recent name change has left Goldman with a mishmash of brands, and turnover has run high since a frenetic pace of product launches left engineers and other employees exhausted. 

There’s even talk that tensions among those who are still there — particularly between the unit’s chief, Peeyush Nahar, and Swati Bhatia, the head of product — are adding to the friction. 

A Goldman spokesperson declined to comment. 

Marcus’ struggles are raising the stakes for Solomon, who built his reputation by bringing ruthless efficiency and cost discipline to Goldman’s investment-banking division. The CEO now finds himself in charge of a money pit at a time when Goldman’s other businesses are staring at a roughly 50% chance of a US recession. Goldman’s stock price has recently flirted with a 52-week low. 

Solomon faced numerous questions from analysts about the consumer business on the company’s second-quarter conference call in mid-July and defended his strategy. 

“I’d highlight that we’ve been building a business from scratch,” Solomon said. “It certainly takes investment, and we’ve been making investment. And we’ve been pretty clear that we have a long-term strategy to create a leading digital platform in the consumer-banking business.”

The superlatives that Solomon chose to describe the business left some room for interpretation. The CEO described the deposit franchise as “really good,” the credit-card business as “interesting” with “interesting technology,” and the robo-advisor as having “some technology.” 

Jason Mikula, a former Marcus employee who now writes a newsletter and consults for financial-technology firms, has described Goldman’s efforts in similarly underwhelming terms. Mikula said there is little doubt the business has failed to live up to its potential. 

“If this weren’t Goldman, and it were a greenfield startup that had a lending product, a credit card, and savings, you’d say, ‘They are doing great,'” Mikula said, ticking off three of Marcus’ existing products. “In the context of Goldman and with how much money they have spent to build this business, that would change my answer considerably.”

Goldman executives have had grand ambitions for the unit ever since they first set it up in 2015. They organized Marcus as a stand-alone entity, and early employees enjoyed a cloistered existence on a floor of the headquarters’ building arranged to evoke Silicon Valley’s finest. A popcorn machine stood in one corner. Employees met for stand-up huddles. And people were encouraged to wear jeans, a previous no-no. 

As Marcus introduced its first product, an installment loan in 2016, executives held trivia nights and other events to create buzz. Excitement picked up when word leaked in mid-2018 that Goldman was working with Apple on a new credit card. It launched roughly a year later.

When Solomon took over from Lloyd Blankfein as Goldman’s CEO in October 2018, he began a series of reviews into Goldman’s various business units, and Marcus attracted his attention. Over the next few months, he came to view the consumer business as one of many siloed businesses that prevented employees from working together as closely as he felt they should, according to sources. 

Now responsible for the company’s stock price, Solomon chafed at the fact that investors weren’t assigning more value to Marcus’ potential. In June 2019, several months after his own analysts wrote about the “deep-pocket risk,” Solomon let out his frustration during a meeting with a couple hundred Marcus employees that was attended by a CNBC reporter. 

“We started out to try to create a business that would disrupt what’s a big, broad industry,” Solomon said. “Now we’re getting absolutely no credit from anybody else in the investing community about that yet.

“If we were out in Silicon Valley and made 20% of the progress that we’ve made, we would get a lot of credit and people would be throwing money at us to own a piece of this business.”

He finished by saying that “nestled inside little old Goldman Sachs, we’re just going to have to prove it over time.”

As the CEO debated whether to weave Marcus more tightly into the firmament of Goldman’s organization, other leaders, including Harit Talwar, the former Discover executive hired to build the business, and his top lieutenant, Omer Ismail, pushed back. They and others in their camp worried that Marcus would lose the culture that had made it so special in the early years, according to people familiar with their views. Solomon and others eventually overruled them. 

In early January 2020, Goldman announced that Marcus would merge with Goldman’s wealth-management unit. The marriage knotted a high-growth digital business with a mature and largely analog investment-management business, and subjected Marcus to battles over corporate resources and political maneuvering for the first time. 

Everyone is talking about how Peeyush and Swati are fighting all the time.

That October, Solomon took another step away from Marcus’ roots by passing over Talwar, who had decades of consumer experience, to lead the combined division. Instead, he chose Stephanie Cohen, the chief strategy officer, and Tucker York, head of the private-wealth business.  The appointments followed a familiar Goldman pattern of naming experienced executives — “good athletes,” in Goldman-speak — to run a business they had never worked in before.

What they lacked in consumer-banking experience, they also lacked in rapport. At least initially. The two soon developed what would become a familiar town-hall routine. Cohen would cite her relatively young age — she is in her mid-40s — and York’s more advanced years, his early 60s. “It was awkward at first,” said a former executive. “They basically got to the point where Tucker became the butt of Stephanie’s jokes.”

Though they soon learned to work together, their tenure has been marked by a wave of departures and confusing strategy choices, according to those who spoke to Insider. Goldman required a prolonged sprint to deliver the Apple Card on time, and its engineering division never recovered. Already stretched to the breaking point, as Insider first reported last year, engineers got little reprieve when executives used that successful launch as a template to press the accelerator for more product launches. Many of those engineers soon quit. 

Turnover didn’t spare the upper ranks. In March 2021, Ismail left for a fintech startup partially owned by Walmart, bringing the head of consumer partnerships with him. Solomon and Waldron persuaded Talwar, who had largely stepped out of the day-to-day decision-making after being passed over for the larger role, to return to his former post. He stepped down a second time later that year. 

The exits continued into this year. Dustin Cohn, the marketing head hired from Jockey International who helped build the early buzz for the business, left for a job with a real-estate startup. Other executives who helped to start the business, such as Elisabeth Kozack, a managing director, also left. 

A January 2021 memo serves as a microcosm of the turnover. In the internal document obtained by Insider, Cohen and York named six executives to run digital products for the division. The six, they wrote, would work together to deliver the bank’s technology platforms and apps in a way that consistently met customer needs, aligned with the business strategy, and made efficient use of resources. 

By March 2022, all six of those executives had either left the bank or taken on different roles. 

In April and May of this year, seven former Marcus employees joined Ismail’s Walmart-backed startup, Mikula reported in his newsletter. Six of those worked in software engineering. 

“The brain drain makes it very hard,” Richard Crone, a payments consultant at Crone Consulting, said in an interview. “The talent, the expertise, the understanding of all the moving parts is in very short supply.”

To be sure, some of the departures from Marcus reflect a red-hot market for talent on Wall Street. Two of the people Insider spoke to said their decision to leave had nothing to do with the environment there but with outside opportunities that were too good to pass up. 

Goldman hired hundreds of new engineers, and executives moved to replace Talwar, Ismail, and other senior executives. Shortly before Ismail’s sudden departure, Cohen hired Swati Bhatia from Stripe, where she was a payments risk executive, to head the proprietary consumer business. Several months later, Goldman hired Peeyush Nahar, from the ride-hailing app Uber, to lead the overall business. 

The business holds more than $100 billion in deposits, has over $16 billion in card balances or loans, and is on pace to log more than $4 billion in revenue by 2024. Goldman has said that much of the investment cost to build the business has already been spent. 

Many of the new costs contributing to this year’s projected $1.2 billion deficit are a result of increased loss reserves that come with higher loan balances, according to the company.

Nonetheless, sources told Insider, those numbers obscure the turmoil that has led to a string of delayed product rollouts, uneven performance, and a muddled strategy. 

There is no greater sign of those struggles, two of the former executives told Insider, than what they have come to learn about the relationship between Nahar and Bhatia: The two executives don’t like to be in the same room together and often prefer not to speak to one another, they said. 

The dynamic between Nahar and Bhatia, according to others who have worked directly with them, traces back to the events around Ismail’s sudden exit. When he left, Bhatia, who has more than 15 years of consumer payments experience at PayPal and Capital One, felt she should have gotten the larger role, said one of the former executives. 

She was resentful when Goldman hired Nahar for the role, an engineer with no consumer-banking experience who is known to be technically proficient but very introverted, the person said. Now, more than a year later, the talk among former employees briefed on the state of their relationship is that it’s effectively broken. 

Cohen has grown so frustrated with Bhatia, in fact, that Cohen reassigned her head of product to the integration of a recent acquisition. 

“Everyone is talking,” the person said, “about how Peeyush and Swati are fighting all the time.” 

When it comes to the delayed product launches, the most glaring is the Marcus checking account, a core banking product that is expected to sit at the center of Goldman’s digital-banking offering. First discussed at Goldman’s January 2020 investor day and slated for a 2021 rollout, the product has run into multiple delays, sources told Insider. 

Earlier this year, Solomon finally promised that Goldman would offer a checking account before December came to a close. And CNBC later reported that the bank had opened it up to beta testing for employees. 

Other product rollouts also arrived late. In January 2021, Goldman announced that it had won the auction to be General Motors partner on the automaker’s credit card. The press release said it would be ready to sign up customers in September of that year, and integrating the card was on Goldman’s 2021 road map, according to one former executive. The card wasn’t available to the public until 2022. 

One person who spoke to Insider said the delays couldn’t all be attributed to Goldman. Delays from Capital One, which held the partnership with GM before Goldman won a new auction, and on GM’s side also slowed the launch. Tensions flared between Goldman and General Motors managers, according to a person familiar with the situation and another who had been briefed on it. 

“From inside, I can say there were delays from the Marcus side, but that was not the only delay,” one former Goldman executive said. 

Goldman finally launched the card with a website attached to its Marcus platform rather than taking additional time to integrate it with GM’s technology. 

And the much-hyped automated wealth offering, dubbed Marcus Invest, seems like an afterthought. At its inception, it stood at a disadvantage to other similar products in the industry, thanks to a fee that was 40% higher than similar offerings from rivals and underlying exchange-traded funds that can be bought more cheaply in the market. 

The robo-advisor rollout has also suffered from divisional infighting over the back-end technology that will power it, pitting Marcus executives in a battle for resources with their colleagues in the global markets division that owns the technology, according to a former executive with knowledge of the deliberations. 

“The Marcus Invest offering itself feels” like a minimum viable product, wrote Mikula, the former employee and newsletter writer, in a February 2021 teardown of the product. “This shouldn’t necessarily come as a surprise; if you stop to reflect on existing Marcus offerings (lending, savings) — these are simple products with good execution, but they don’t offer much differentiation vs. others in the market beyond having Goldman’s name attached.”

In late June or early July, Goldman reduced the fee for Marcus Invest to bring it in line with the 25 basis point fee at rivals like Betterment, and removed the minimum investment amount, to better attract retail customers. 

Sometimes, it feels like there is not a single clear strategy at play.

One product that was distinct and came to define Goldman’s consumer-banking ambitions is the Apple Card. Launched during Solomon’s first year as CEO, the partnership continues to lend credibility to Goldman’s efforts. 

But even the Apple Card has presented challenges for Marcus’ strategy, sources told Insider. In 2019, a software developer criticized the card’s underwriting decision when it rejected his wife for a credit line increase, despite, he said, her higher credit score. 

“My wife and I filed joint tax returns, live in a community-property state, and have been married for a long time,” the executive wrote on Twitter. “Yet Apple’s black box algorithm thinks I deserve 20x the credit limit she does.” 

New York State regulators launched an investigation, later finding no evidence of bias. 

Earlier this year, Apple announced plans to move underwriting for a new BNPL — or “buy now, pay later” — product in-house, dealing a blow to Goldman’s ambitions to maintain a tight partnership with the consumer-tech behemoth. 

And earlier this month, Goldman said it was cooperating with regulators who were investigating the firm’s “credit card account management practices,” such as applying refunds, resolving billing errors, and reporting customer information to credit bureaus. 

Apple also extracted concessions from Goldman that have made it more difficult to forge a unified strategy, sources said. During discussions over the Apple Card, Goldman executives pushed to use the “Marcus by Goldman Sachs” brand on the back of the card, one of the sources said. Apple balked, insisting on “Goldman Sachs Bank USA,” the person said. Goldman agreed, and the Goldman Sachs Bank USA brand appears on the back of every Apple Card and the documentation for the product. 

Since then, Goldman has pressed successfully to use its Marcus by Goldman Sachs brand with General Motors and JetBlue, with whom Goldman offers a buy now, pay later option to purchase plane tickets. The result is various brand names standing behind the company’s card offerings. 

Goldman has even changed the name of the business. In October 2021, Cohen told the crowd at the Money 2020 conference that the bank had interviewed thousands of customers and learned that they wanted to be “closer to the brand,” necessitating a change in the name to “Goldman Sachs Marcus.” 

Nearly a year later, Goldman has yet to introduce the new name and the firm’s website still uses the “Marcus by Goldman Sachs” branding. 

On Goldman’s July earnings call, Solomon seemed to indicate that Goldman knew it needed to streamline its strategy. In answer to an analyst’s question about new partnerships, the CEO said 2022 would be a year when Goldman would catch its breath and look to tie its various products together more tightly.  

“Our intention at the moment is to be focused on integrating these successfully and making sure we execute at a very high level,” the CEO said. “Certainly, as we get out to 2023 and 2024, we will be more open to other partnerships and other meaningful things going forward.”

Despite the turmoil, Goldman’s consumer-banking business touts an impressive number of customers: 14 million. Most five-year-old digital-banking startups would be thrilled with those numbers. “Show me any organization which has, five years after launch, so many products and partnerships that are doing well, I would find it hard to find one,” said another former executive. 

But scratch the surface, according to two former executives, and the numbers aren’t as impressive as they may seem. 

Apple has allowed Goldman to reach millions of customers instantly through its partner’s devices, and card clients still account for the majority of the unit’s customers. Roughly “high-single digits” of the 14 million customers are Apple Card customers, according to one former executive. Crone, the payments consultant, estimates the Apple Card has 7 million active users, which it defines as those making two or more transactions a month. 

Another 3 million are customers of the General Motors card, according to disclosures made by the automaker. GreenSky, a BNPL company Goldman bought last year, as well as partnerships with JetBlue and Amazon, add more users. 

Goldman doesn’t disclose how many customers use more than one of its products, but insiders say it isn’t that many. That means that Marcus, Goldman’s core digital bank, has just a few million of its own. Compare that to the 70 million or so claimed by the payment app Venmo or the 23 million that the trading app Robinhood had at the end of June. 

According to one former executive, Goldman’s figure may still include those people who have simply downloaded the Marcus app and logged on to Marcus Insights, the successor to the Clarity Money app that analyzes a consumer’s financial picture. 

Add it all up, and the picture of Marcus that emerges is one of a business that’s in disarray. 

“Sometimes,” Mikula said, “it feels like there is not a single clear strategy at play.”