Download to read offline
More Related Content
Similar to WHAT IS A DATA LAKE? Know DATA LAKES & SALES ECOSYSTEM
![Get a Pinterest account in[ pvatopservice ]sell regular every day.docx](https://cdn.slidesharecdn.com/ss_thumbnails/getapinterestaccountinpvatopservicesellregulareveryday-251030182004-67b5ef9f-thumbnail.jpg?width=640&height=640&fit=bounds)
WHAT IS A DATA LAKE? Know DATA LAKES & SALES ECOSYSTEM
- 2. INTRODUCTION The volume, variety,velocity and veracity of big data are getting increasingly complex each passing day. The way the data is stored, processed, managed and shared with decision-makers is getting impacted by this complexity and to tackle the same, a revolutionary approach to data management has come into picture. A data lake. ©Denave
- 3.
- 4. ©Denave As the namesuggests, data lake is a large reservoir of data – structured or unstructured, fed through disparate channels. The data is fed through channels in an ad-hoc manner into these data lakes, however, owing to the predefined set of rules or schema, correlation between the database is established automatically to help with the extraction of meaningful information. It provides high level of flexibility in terms of interaction with and leverage of the data. In general, data lakes are used to store data when you’ve a constant stream of unstructured data coming into, such as, web interactions, product logs, IoT sensors, app usage etc. Simply put, along with the on-premises data, it is the real-time data which fills up the data lake, upon which are then used the principles of machine learning and analytics to make it relational. The global data lakes market is expected to grow at a CAGR of approximately 28% during the forecast period 2017-2023, to touch $12.01 billion by 2024 & $+14.01 billion by 2026. 1,2&3
- 5. ©Denave DATA LAKES & SALESECOSYSTEM UNDERSTANDING THE CORRELATION
- 6. Sales is nomore just about the product alone. Rather, it is more about getting connected to the customer – at a deeper level. In order to do so, organisations are becoming data-driven in every sense and they rely heavily on agile technologies to help them with the analysis and management of data. With a large wealth of customer data at their disposal, it is that data analysis backed sales action which acts as a game-changer for organisations. To get the differentiating edge, firms need to use the customer data in the best possible manner to make hyper-impactful outreach and better sales interactions. Following elucidates some of the challenges of sales ecosystem which are solved by Data lakes: Need for the Data in Native Format It may be difficult to believe but since the amount of data being generated by enterprises is extremely huge, a major portion of that data is discarded. Remaining small portion is then stored in the data warehouse – for a few years. This happens owing to the storage capacity, structure restriction, associated costs etc. and most of the time it happens because enterprises don’t know what to be done with that data, esp. machine-generated or historical data. Hence, it is dumped away, thus putting a limitation on the extent of analytics application that can happen. With data lakes, enterprises save the data without fretting about the structure, intended use etc. Simply put, you may not know why you are saving the data, but you’d still do so with the thought that you may need that data someday – thus, getting as much data as possible in its native format. ©Denave
- 7. With data lakes,enterprises save the data without fretting about the structure, intended use etc. Simply put, you may not know why you are saving the data, but you’d still do so with the thought that you may need that data someday – thus, getting as much data as possible in its native format. Quashing Data Silos With data, the case is such that the number of teams or departments you have, that much variety of data would be there. That database is not centralised and instead remain in silos because it turns out to be expensive as well as time consuming to share that data with one another. Consider the case where a department may need the data from another department – their requirement would be a specific segment of data in a particular format, Therefore, the department which owns the data will be conducting all ETL exercises to be able to extract and package the data in line with the requirement. Extra work equals to extra time and also extra expenses and hence the delivery team would want to shy away from getting any data request in the first place or to delay and excuse any request which has already been made, resulting in data silos. It is like having an immense wealth with you but not being able to use it owing to the labours it is going to take to spend it. This issue is tackled head on with data lakes because there the data ingestion is almost frictionless since it accepts data without any processing, thus allows for a deeper data leverage to all with a centralised and transparent access process. Data ownership doesn’t remain a barrier any more with data lakes. ©Denave
- 8.
- 9. The acceptance ofneed for data lakes itself makes visible a lot of benefits which salespeople can accrue, let’s dig deeper and see what all as an organisation are you set to gain if your database strategies include leveraging data lakes: Say good-bye to silos and fragmentation With data lakes, you get a unified view of everything which comprises the customer experience – all the data from all the platforms, departments, teams, delivery channels. Better preparedness for the customer journey Since you’ve better knowledge about the customer buying cycle, the high or low points of his journey, quite naturally the decision-making process is much more impactful than before. High yielding campaigns You’re better equipped than before for generating and assessing campaigns. The agility and autonomy are rendered by the tools which not only helps you to measure but also to optimise your marketing investments in real time. Power to predict Sitting on the historical data which you’d have otherwise dumped or would have never been able to analyse, provides you with the power to analyse and establish predictive models for customer behaviour. ©Denave
- 10. All touch pointcontrol You’re also able to trace the journey and behaviour specific to each customer touchpoint which allows you to rank the point of conversions accurately. Accordingly, you can adjust your focus on the touchpoints. Fertile environment for new tools or methodologies development Since you own the storage of immense amount of data, you can generate your own technological environment and a service-oriented architecture which can make the development of new tools possible. Intelligent and faster decisions With dynamic dashboards which work upon varied kinds of data accumulated from myriad number of sources – external or internal, you get an unmatched business intelligence right at your tips and thus, decision-making in real-time becomes possible. If we put together all the benefits in one basket, the highlight of data lake adoption would be the provision of an agile 360-degree view data-driven marketing operations and a faster (and intelligent) response system to any business need. These will automatically translate into improved customer interactions, enhanced R&D innovation possibilities and augmented operational efficiencies. ©Denave
- 11.
- 12. Getting convinced ofthe immense advantages of data lakes is one thing but understanding where it fits in your business needs and vision is another thing and probably the more important one. A quick assessment of the latter will help in getting the optimal resolution for your business: Understand your data type What is the type of data which forms the majority of your data input/ accumulation? If it is structured data in generality – traditional, tabular format – data being generated by HR systems, traditional CRMs etc. then data warehouse is a good option. Data lakes are beneficial in case if the data you have is unstructured or semi- structured and it’s getting generated constantly (like a person crosses an in-store ad and the data is captured by the intelligent display unit to get records of user views and behaviour). In such a constant stream of data, hundreds of terabytes are generated and occupy the storage in a matter of time. In short, your structured data may be more suited for RDBMS while the sensor or SAAS based data would be more suited for data lakes. If you would use a data lake for the prior stated data variety then the costs would be exorbitant and unrequired. ©Denave
- 13. Understand your dataexpectation If the post storage needs from the data are fixed and you need it for specific use cases, then you’d structure the data in advance for the requite processing. For such kind of expectation from the data, a data warehouse is best suited, however, structuring the data has its own costs attached and it certainly limits your ability to repurpose the data for any other use later on. With data lakes, comes in the flexibility. You can store now and decide later what you really want to extract out of that data. For a set of predefined reports, you may not bother having a data lake but to have the freedom to analyse and accordingly scale up or innovate, data lakes would be the best bet. Understand your resources While DB technicians or IT team would be able to get a database system in place and then business analysts can leverage self-service to extract reports out of it, data lakes would require a different level of expertise. Skilled big data engineers and data scientists would have to be invested in along with sophisticated tools to form a data puddle and then a data pond and finally a data lake. Analyse your data acquisition process If the data acquisition is happening from multiple sources and is a complex process in itself -it would mean that you’d be spending a large overhead on ETL techniques in order to render that data suitable for data warehouse. If you’re constantly spending a fortune because of a complex data acquisition and thus a cost intensive process, then it might be time to switch to data lake. Once you take the plunge into data lakes based on your requirements, remember the governance part as well and involve the legal & privacy team for required due diligence. ©Denave
- 14.
- 15. Getting the rightplatform and then getting the data loaded onto it are the first steps but the real job begins post that. The interface plays a critical role in adoption and usage of data lake. The scope of data lake is huge enough that the normal IT function, which is adequate to manage a data warehouse, can’t be expected to cater to such a wide variety of data as well as a huge community of users. Therefore, self service becomes an essential feature for the data lake interface. It carries these two integral aspects: • To ensure the access grant in accordance with the level of expertise • To ensure that people are able to find the right data that they are looking for For example – a business analysts would need a cooked form of data and a completely raw data set would not make sense to him/ her, though getting the same unprocessed data would be necessary for a data scientist. Hence, the interface will have to have bifurcated zones for different data requirements of divergent set of audience, such as, partially processed data, raw data etc. ©Denave
- 16. CONCLUSION The whole processof getting the infrastructure in place for the data lake, organising the data lake (creating zones for various user communities as per their expertise level), putting up the catalogue of data assets to enable self-service and eventually, opening it for the users – this complete journey is like a success roadmap for data lakes. Such kind of centralisation of data allows for multiple forms of analysis such as analysing the conversion funnel for improvement areas, creation of a precise recommendation engine etc. – everything facilitating engaging customer experiences and explosive business growth. ©Denave
- 17.
- 18. For more salesinsights, visit www.denave.com/resources ©Denave