The Global Video Analytics Market was valued at USD 1.528 billion in 2020 and expected to reach USD 4.142 billion by 2026 and grow at a CAGR of 24.5% over the forecast period (2021−2026) .
At the current stage, the main consumers are in the retail, manufacturing, and transport sectors. However, the financial industry is in no rush to increase its momentum, despite the huge number of tasks in the security and service optimization fields that could be profitably resolved by introducing video analytics. In this article we explore what technologies a modern bank absolutely can not do without, and what exactly is slowing their introduction.
The modern bank consists much less of just its branch building than all of its digital operations. Through biometrics, truly modern levels of efficiency and security can be achieved.
Biometric authorization technologies have been gradually replacing document-based identity verification, helping simplify many processes for a financial institution’s clients and employees alike. Upon the first visit to the bank, the customer submits their biometrics, that identity is linked to certain documents and other data, and thereafter the need to present an ID and come to the branch in person practically disappears. Opening a new account or getting a new card, applying for a loan, confirming a transaction: all this can be done on a smartphone, using only your face for confirmation. The level of personal data protection corresponds to industry security standards, and for additional confidence it is possible to configure up to three identification factors: password, SMS confirmation, and facial biometrics.
In moving to biometrics, banks have a lot of work to do: they need to organize data collection directly in the offices and through mobile applications while building an ecosystem for reception and processing. But in practice, these investments more than pay for themselves by a smaller number of branches needing servicing and additional profits from the inflow of customers from hard-to-reach places such as rural areas.
Payments without a card or even a smartphone
The rejection of physical cards is already on a massive scale, people transferring the card to their smartphone or immediately ordering virtual cards instead of plastic. The next step on this path may be payment by face. The technology for this is quite functional, not only with online purchases on marketplaces, but also at ordinary grocery stores, shopping centers, and gas stations. It works well enough to install a device at the point of sale capable of reading biometric data. The system works through the bank: the client links their biometrics to the card in the branch office, and payment by selfie becomes available. For buyers, it’s the comfort and speed of service. For financial companies, it’s the expansion of potential loyal customers and improvement of their image.
This idea has already been implemented by many banks around the world: clients submit biometric data and attach it to their profile. A bank card can be linked here. The system maintains a high level of payment security: it is impossible to fool the terminal using someone else’s photo or smartphone video, thanks to the protective algorithm against such forgeries. In addition, the payment process takes place three times faster than via a card or smartphone, because it is not necessary to confirm the transaction with a pin code or SMS numbers. The user’s biometric verification is also available for translation in the app.
Improving customer service quality
One of the most commonly heard scenarios is line optimization. Here the principle is the same as in the retail one: video analytics allows for collecting statistics on department loads, determining peak hours at each point, and recording the most common visitors’ requests. By analyzing these data, it is possible to optimize the managers’ and consultants’ work and efficiently redistribute them to the hottest areas.
The second most common scenario is to increase servicing speed in offices. When a visitor enters a branch, the system automatically identifies them, referring to the internal database and checking the latest transactions to suggest the most suitable manager. This is faster and more efficient than taking a number on a paper ticket where the client has to choose the purpose of their visit on the terminal. Despite the general trend of reducing lines at banks, the problem remains, because the servicing process itself is lengthening. It currently fluctuates in the range of between 10 minutes to an hour.
The financial industry now needs solutions that not only record what is happening, but also react quickly to dangerous situations. And ideally, prevent them. This is particularly relevant to safety and health issues. If, for example, an intruder enters the branch with a weapon, it is necessary to react to them not at the moment when they draw the pistol, but rather immediately, at the entrance. The system sends a danger signal to law enforcement, detecting suspicion on certain grounds, or in the case of identification of a person on a particular list (wanted or in the database of unscrupulous borrowers).
To do this, lists formed by one bank, common lists for several organizations, or databases provided by law enforcement agencies can be used. The neural network will recognize the face on the camera video if it matches the data from the database and instantly send a signal to the responsible employee.
Situations where the use of video analytics can significantly reduce risks also include crowds near ATMs or the prolonged stay of subjects near the ATM without a clear reason. Today, even such scenarios are possible when the system identifies a user’s attempts to deliberately hide their face (simultaneous use of a mask, hat, hood, etc.) and thereby prevents them from using the ATM. In such cases, hijabs or medical masks will not be signs of a criminal, this concerns situations where a person is purposefully bundled up so that 10−15% of the face is exposed in the eye area. It is important that the final decision in this situation is made by the operator, who receives a notification from the security system, and can take timely action to catch the attacker and prevent an act of robbery or vandalism.
Another problem effectively dealt with by video analytics are loans taken out on someone else’s passport. This is unfortunately still quite relevant for microfinance organizations. Usually, fraudsters use a copy of someone else’s passport, stolen or lost original, paste their photo there, and take out a loan for someone else. The introduction of video analytics and biometrics technologies helps to counter such situations. If a client of the bank has submitted their biometrics, the system will immediately recognize the wrong passport and fake photo. For online loans, this mechanism also provides reliable protection. After a foiled attempt, the fraudster is blacklisted, and if the bank is configured with the appropriate option, the information can go directly to law enforcement.
Statistics say that company employees spend on average up to 14 hours per week not on their direct duties, but on smoke and coffee breaks. We can also include personal phone conversations, social networks, reading news, playing solitaire on the computer when the boss is away. Resting is necessary, of course, but its abuse accumulates in significant figures of losses for the company. At the same time, more than 90% of HR managers have no ability to meaningfully influence the situation. So, we have video analytics. It helps control the employee in the workplace, record the time of their arrival and departure, and movement within the office.
But these are not the only things that matter to the banking sector. At the forefront, there are security issues, such as crossing boundaries in violation of access rights. There is also the issue of abuse of official powers and improper operations.
The use of video analytics, especially in conjunction with biometrics, can minimize illegal actions and increase employee responsibility and work discipline.
Complications in introducing video analytics in banks
The introduction of any system in banks requires long and expensive cycles, due to compatibility, compliance, and security issues being critical to financial organizations. This also applies to video analytics, as it implies changing fundamental banking processes. Not all organizations are ready for such long and complex integration, so the introduction of new technologies has been delayed.
Despite the popular and well-founded opinion about the high technology of banks, there is a huge gap between the conditional top ten banks and all the rest. They are all at various technological levels. Giants innovate slowly but surely, but small banks often don’t even have any automated compliance, with everything being done manually. They have no need to discuss the technologies of video analytics and biometrics.
Last but not least, personal data is another challenge. People don’t trust financial institutions enough to actively give their biometric info to banks. So banks aren’t very willing to introduce biometrics: they don’t see the demand. Until the issue of consumer confidence has been resolved, they will continue doubting.
 Video Analytics Market — Growth, Trends, COVID-19 Impact, 2021−2026) by Mordor Intelligence