91̽»¨

March's retail sales may have plateaued, but the 1.9% quarter-on-quarter increase in sales volumes in Q1 is a promising indicator that the downturn from the latter half of 2024 has ended. 

Looking ahead, we anticipate retail sales to steadily improve as the year progresses. This is underpinned by several factors. Firstly, the imminent surge in households' real disposable incomes, driven by a reduction in inflation and the implementation of tax cuts from April. This should boost consumer spending and, consequently, retail sales volumes. Additionally, a continued uptrend in consumer confidence is expected to ensure that the majority of this newfound income will circulate back into the economy.

Inflation within the retail sector remains notably subdued, with retail goods prices increasing by just 2.2% year-on-year in March, marking the most gradual increase since early 2021. This trend suggests that robust nominal spending will increasingly be reflected in actual sales volumes. 

Lastly, retail sales volumes linger approximately 2% beneath their pre-pandemic figures, a residual effect of consumers’ substantial spending on goods during the pandemic. However, after two years of moderated spending on retail items, there is an emerging necessity for households to begin replenishing goods acquired during that period. The revival of the housing market and an increase in transactions are projected to stimulate demand for household goods, adding another layer of support for retail sales. Overall, it’s a much more positive outlook off the back of Q1 as we enter the second quarter.

As we examine the key themes across retail in 2024, there’s no getting away from the prominence that crime has played in the industry since cost-of-living pressures set in. In this edition of our retail outlook, we’ll be focusing on current crime rate data for the sector, and data-led solutions retailers could implement to tackle this issue. 

Rise in retail crime

Crime has proved a significant issue for retailers in recent years. Growing economic pressures have created a challenging living situation for many, and unfortunately this has resulted in a direct spike in crime since Covid induced a lull in incidents. Additionally, there has been an uptick in organized crime activities facilitated through social media platforms. A notable example is the now infamous Oxford Street robbery campaign that unfolded in August 2023, orchestrated via TikTok. 

Shoplifting offences for the year ending December 2023 rose by over a third year on year (37%) and are at their highest level in 20 years reaching 430,104 offences compared with 315,040 in 2022. 

Bringing these numbers to life, retailers have started to share their own crime data in a bid to drive awareness. For example, Co-op reported approximately 1,000 cases of crime in its stores every day in the first half of 2023, including shoplifting and antisocial behaviour.

Unfortunately, there’s an underlying sentiment in the industry that this problem isn’t going away. Historically there’s been a lack of drive from law enforcement to tackle this issue, which has discouraged retailers from reporting crimes. A freedom of information request made by Co-op in 2023 revealed that police did not respond to 71% of serious retail crimes. Following this, the British Retail Consortium (BRC) found that only 36% of violence and crime incidents experienced by retailers were reported to police. 

However, 2024 has seen a positive shift, with policymakers beginning to acknowledge the severity of the issue. In a significant move, the in April that assaulting a retail worker would now be classified as a standalone criminal offence. While this marks a crucial step forward and is expected to increase the reporting numbers for retail crime, the industry remains committed to investment in crime prevention measures. The BRC reported that the last year alone. With criminal offences still on the rise, retailers must remain vigilant in addressing this issue internally to curb shrinkage and safeguard profit margins.

How to use data to tackle shoplifting?

Retailers are harnessing the power of data and facial recognition technology to tackle their growing shoplifting problem, a solution reportedly endorsed by the UK policing minister. Many retailers already have CCTV in place and are performing reviews of this data after incidents occur. However, this is retrospective and manual. There is a movement to become more proactive with this information by leveraging CCTV data in partnership with facial recognition technology. This can help automate the identification of thieves as well as enable the creation of watchlists.

As great as these technologies sound, there are five key factors to consider before leveraging CCTV data and investing in facial recognition models for crime prevention. 

1. Purpose-driven approach: retailers must prioritise defining a clear use case or problem statement for their data analytics initiatives. They should take ownership of their strategy by asking critical questions about the intended outcomes, data sources, data protection compliance, system integration, and potential risks throughout the data lifecycle.
2. Data Protection Impact Assessment (DPIA): it's essential to involve an independent data protection officer in evaluating and mitigating risks related to data protection laws. Conducting a DPIA helps identify and address privacy concerns associated with data analytics efforts.
3. Handling sensitive data: recognise that facial recognition technology relies on biometric data, which is highly sensitive and subject to strict legislation. Establish meticulous guidelines for managing biometric data, including criteria for watchlist inclusion and removal to comply with privacy regulations.
4. Ensuring model accuracy: acknowledge that the effectiveness of facial recognition models hinges on the quality of the underlying data. Be wary of anyone promising 100% accuracy in their solutions and be sure to question how statistical accuracy is ensured. Utilise resources like the National Institute of Standards and Technology (NIST) for independent evaluation and validation of model performance.
5. Human-centric approach: while CCTV data and facial recognition technology are valuable tools in combating retail crime, it's crucial not to overlook the human element. Maintain a balance between automated solutions and human oversight to ensure the ethical and effective application of facial recognition technology in real-world scenarios.

Data’s role in detecting employee fraud

Research by found that the value of retail crime was forecast at £7.9bn in 2023. Alarmingly, the report revealed that 40% of the total value of theft is attributed to employees. Although there are generally fewer instances of employee crime compared to those committed via external actors, internal crime tends to involve more frequent levels of theft and larger volumes of goods. Average losses from a dishonest employee are four times the amount of those of a shoplifter.

What’s driving this spike in employee crime in retail, particularly in Distribution Centres (DCs)? Retailers attribute the current economic climate as a key driver, with 51% of retailers noting greater reliance on temporary staff due to labour shortages and 48% referencing financial need from the cost-of-living crisis.

Retailers can harness data analytics to detect fraud and theft by identifying irregular purchasing patterns by staff and tracking inventory levels. For example, analytics based on point-of-sale data can help identify where products are frequently marked down by a staff member who is colluding with an external actor to commit fraud by improperly discounting goods. 

Data analytics can also be used to identify discrepancies in inventory levels and the inventory recorded in systems. However, the analysis around inventory can only be as good as the quality of an organisation’s data. So, when pursuing analytics, it is critical to have proper data governance and management practices in place. 

Top five ways retailers can harness data analytics to prevent employee theft

1. Anomaly detection: identify unusual patterns or anomalies by leveraging sales, inventory, and transaction data. By analysing historical data, retailers can establish a baseline of normal behaviour and then flag any deviations from this norm. Sudden spikes or drops in sales or inventory levels can indicate potential theft or fraudulent activities.
2. Employee behaviour analysis: monitor employee behaviour through the tracking and analysis of activities such as sales performance, transaction history, and access to sensitive areas. This can help detect suspicious behaviour like excessive voids or refunds.
3. Predictive modelling: employ predictive modelling techniques to forecast potential instances of employee theft based on historical data and risk factors of known employee theft. This allows for proactive measures to prevent theft before it occurs.
4. Video analytics: enhance video surveillance systems with AI-trained models to analyse footage in real-time or post-event, detecting suspicious behaviour like unauthorised access or instances of theft caught on camera.
5. Inventory management optimisation: optimise inventory management processes using analytics to identify discrepancies between expected and actual inventory levels. Closely monitor inventory movements and conduct regular audits to address instances of theft effectively.

Overall, by harnessing the power of data effectively, retailers can enhance their ability to detect and prevent employee theft, ultimately minimising losses and improving overall operational efficiency.

If you’d like to discuss any of the issues raised in this article, please contact our experts: