Tariffs, viral trend cycles, and fragmented systems have turned inventory planning into one of the most stressful jobs in retail.
DOSS surveyed 661 retail and consumer goods professionals across supply chain, procurement, forecasting, and finance, and what they admitted should raise eyebrows across the industry:
- three in four retail professionals (75%) have either lost sleep over an inventory decision or come close to it, losing an average of three hours of sleep;
- Seasonal or trend-driven SKUs and perishable goods are tied as the most stressful inventory categories to manage among retail professionals, each cited by 33% of respondents;
- one in two retail industry organizations (50%) delayed or canceled purchase orders due to tariff uncertainty in 2026;
- More than half of retail professionals (51%) say they sometimes proceed with demand forecasts they know are unreliable, simply because no better alternative exists;
- 64% of retail industry organizations have not deployed AI or machine learning for inventory management, and among those with no plans to adopt, 59% say it’s just not a current priority;
- Among the one in four retail industry organizations left holding excess inventory after a viral trend faded, TikTok was the responsible platform 60% of the time, followed by Instagram (13%) and YouTube (7%).
You can explore the full study here.
David Appel, VP of Marketing for DOSS, spoke with Retail Insider about the issue.
Question: How are tariffs and ongoing uncertainty reshaping inventory planning strategies in 2026?
Answer: Tariffs have thrown many into disarray. In the past, the formula for success seemed clear. It was find trusted foreign suppliers, create lasting partnerships, focus on costs, and keep the overhead low. Yet, once tariffs are changed, everything that went into these calculations instantly turns to dust. Half of the businesses surveyed deferred or canceled POs during this year alone due to tariff unpredictability. It is not surprising that more savvy companies have started to look at other geographic sources at the risk of higher costs. Moreover, they have started to add safety nets and plan according to scenarios rather than relying solely on the numbers. It may decrease efficiency temporarily, yet it will help maintain stability, which has never been easy to ensure.
Q: Why are so many retail professionals still relying on demand forecasts they know are unreliable, and what needs to change?
A: In some ways, it has become an issue of not yet knowing what better looks like. Forecasting processes become deeply ingrained in planning, budgeting, and other activities within an organization. As far as the human aspect goes, it is hard to recognize that what worked in the past may become unreliable and inefficient. According to our survey, 51% of professionals continue to use their current forecasting methods, despite their unreliability, simply because they do not have anything else. Also, 16% of respondents do not track forecasting errors at all. However, you cannot fix a problem you do not even measure. Instead, one should stop regarding it as a fixed point and try to evolve constantly by using rolling forecast processes that are adjusted depending on the situation. Ideally, assumptions and forecasts should be revised on a weekly or even daily basis during periods of volatility. One mistake with a large retailer can severely impact a business.
Q: What makes seasonal, trend-driven, and perishable inventory particularly stressful to manage today?
A: These categories are among the most challenging to deal with because they have one common trait: time pressure. A seasonal item, such as a winter jacket that remains unsold after February, does not just mean lower profit margins. It means an additional burden for the following year. Unfortunately, the matter is complicated further by growing and increasingly unpredictable demand drivers that make it difficult to forecast accurately. Not so long ago, one could make a fairly reliable prediction by analyzing historical data, seasonal weather forecasts, promotional calendars, and other factors. Today, one has to consider tariff-driven costs, unpredictable social media-fueled demand surges, and increasingly volatile consumer behaviour. The stress experienced is not limited to logistics. It is associated with understanding that the tool set that they have was not designed for such conditions. Seasonal and trend-driven SKUs and perishables turned out to be equally stressful, and in my opinion, there is a reason behind it. The planning cycle became shorter, and there is less space for error.
Q: Why has AI and machine learning adoption in inventory management remained so low despite clear pain points?
A: AI technology was oversold to the industry in terms of its potential benefits. Many executives tend to believe that the adoption of AI would instantly solve all of their forecasting challenges, and it would be easy to integrate into existing processes. However, it requires proper underlying infrastructure and consistent, high-quality data that many organizations still lack. According to our survey, nearly one in five inventory teams need access to four or more data silos to evaluate their stock health. Thus, trying to use AI in such a case would only lead to confusion, which will only be automated. 59% of non-adopters in our survey state that implementing artificial intelligence is not one of their priorities at the moment. While some of it is understandable considering operational difficulties and priorities, I believe that another reason for it is uncertainty. Companies that were able to benefit from AI have done several things. First of all, they invested heavily in data infrastructure, and secondly, they viewed it not as a quick solution but rather as a tool for continuous improvement.
Q: How are viral platforms like TikTok influencing inventory risk, and what can retailers do to avoid being stuck with excess stock?
A: What TikTok brought that many retailers did not anticipate is demand randomness. Prior to it, most demand spikes had clear triggers that could be anticipated. Good holiday season, product placement in a movie or commercial, and endorsement. Today, TikTok generates demand surges that are not predictable and happen much faster. Our survey showed that 60% of viral inventory overstocks were generated by TikTok. Moreover, they take longer to recover compared to conventional spikes. 74% of affected companies took up to six months to clear the surplus. Unfortunately, many of them tried to respond to the surge by ordering additional stocks, which proved to be a huge mistake. By the time it arrived, the hype had already passed. To avoid overstocking, companies are looking at other approaches, such as more flexible supplier relationships, focusing on domestic or nearshore production, and limiting stock increases until there is solid proof of sustainability.
More from Retail Insider:
