0:02 So, you're considering a career in data
0:04 analytics, but have little idea of what
0:05 your day-to-day would look like and
0:07 whether you're going to be replaced by
0:09 AI? Well, this might be the video for
0:12 you. If you're new here, hi, I'm Ingret.
0:14 I share my career journey and life
0:15 online. And for the last 7 years, I have
0:17 been working as a data analyst in
0:19 London. Since this career path is
0:20 evolving every year, I wanted to chat
0:22 through some of the data analytics roles
0:25 still available in 2025 and what I
0:28 personally do as a lead data analyst.
0:30 We're also going to graze over the
0:32 salaries and the scary AI taking away
0:40 The role at its core is to use data to
0:43 solve business problems. You do that by
0:44 talking to your stakeholders to
0:46 understand the real problems that
0:48 they're facing and real questions that
0:50 they're trying to get answers to. You
0:52 then identify the types of data that
0:54 you're going to need to answer those
0:56 questions. where is it stored and
0:58 whether it's even reliable for you to
1:01 use. You then code 90% of the time in
1:04 SQL, sometimes in Python to clean, join,
1:06 aggregate, and manipulate those
1:08 different data sets. Next, you analyze
1:10 the data to spot trends and patterns.
1:12 And after all of this analysis, you
1:15 visualize the most important insights or
1:17 metrics through data visualizations,
1:20 dashboards, reports, or slide decks. And
1:22 at the end of it all, you present your
1:24 findings and recommendations to your
1:25 stakeholders who asked you that question
1:27 in the first place. What I would love
1:29 for you to take away from all of this is
1:31 that the key here is not the coding.
1:33 It's to be able to connect the numbers
1:34 and the trends and the patterns that
1:35 you're seeing back to the problem that
1:37 your stakeholders are facing and help
1:39 them solve that problem through data.
1:40 That's what differentiates a query
1:42 monkey from a really good
1:44 businessoriented data analyst. Remember,
1:46 coding is just a tool to help you answer
1:51 One of the most common data analytics
1:54 roles is a hybrid role between building
1:56 automated dashboards and interactive
1:59 reports as well as answering ad hoc
2:01 business questions. That role might just
2:03 be called a data analyst. You might be
2:05 called a business analyst. And if it's a
2:07 lot more reporting and dashboard
2:09 focused, you might be called a business
2:12 intelligence analyst. As this type of
2:13 data analyst, you're basically helping
2:15 the business to make sense of all of the
2:17 data that is available to them. So some
2:19 of the role will involve building
2:21 interactive dashboards and automating
2:23 some of the reports so that people in
2:26 the business can selfserve the data in
2:28 an easy interactive and visual way.
2:30 However, it's not just about pulling the
2:32 data and building a dashboard. It's
2:33 talking to your stakeholders to define
2:35 the requirements for the dashboard. You
2:37 agree how you're going to define the
2:40 metrics such as an active user or a
2:43 churned user. A will also involve making
2:44 sure that people in the business have a
2:46 shared understanding of that. And then
2:48 in addition, you're also likely to get
2:50 some ad hoc requests. Any sort of
2:51 question that your business stakeholders
2:53 has, such as, "We've launched a referral
2:56 program. Which city did it perform best
2:58 in?" And this is your opportunity to dig
2:59 a little bit deeper into the data on
3:01 some specific business question and then
3:03 create a collection of data
3:05 visualizations which will tell a story
3:07 to answer that question that you are
3:08 being asked by your business
3:09 stakeholders. This role in most of the
3:11 data analytics roles, you will be
3:13 working with SQL every single day. You
3:15 will be using it to clean, manipulate,
3:16 aggregate the data set. You might be
3:19 creating new data sources and ensuring
3:21 data quality. This role is a very
3:23 generic data analytics role. It is a
3:25 great starting role for any data analyst
3:27 as it helps you build really good
3:29 foundations in data analysis, in data
3:31 visualization and reporting and you get
3:34 introduced in how to use data in solving
3:40 Another common branch of data analytics
3:42 is data insights analyst. And if the
3:44 role of a previously discussed data
3:46 analyst is to build out reporting and
3:49 the metrics to help business understand
3:51 its performance, the role of an insights
3:53 analyst is to take those metrics one
3:55 step further and understanding what
3:57 drives the change in the metric or user
3:59 behavior. Now, this is quite a different
4:02 role because it relies more heavily on
4:04 having good business acumen, problem
4:05 solving, storytelling, and presentation
4:07 skills. And we're going to deep dive
4:08 into that in just a second because
4:12 that's a role that I specialize in.
4:15 Another area which is on the rise is
4:18 product analytics. The product analyst
4:19 almost combines the role of the two
4:21 discussed previously. However, they are
4:24 embedded to a specific product. And when
4:26 I say product, I don't mean the physical
4:28 product. I mean a digital experience or
4:30 a service that a company offers. Could
4:32 be an entire website or just a checkout
4:36 flow or a mobile apps landing page or an
4:39 onboarding flow for new users. They're
4:41 also embedded in a cross functional
4:43 product squad and that's a bunch of
4:45 people with different skills trying to
4:47 build that experience and bring it to
4:49 life. Because of that setup, problem
4:52 solving, product thinking, communication
4:55 are extremely key parts of a product
4:57 analyst role and you will be responsible
4:59 for measuring and understanding the
5:01 product performance, user behavior and
5:03 giving recommendations on how to improve
5:05 that experience. Now, in some companies,
5:07 it gets a bit more technical and some of
5:10 the product analysts are responsible for
5:12 experimentation. However, that part of
5:14 the role is a little bit more technical
5:17 and in my opinion overlaps with product
5:19 data science and that's a field that I
5:22 want to ultimately get into. So, okay,
5:24 Ingred, but what do you do as a data
5:30 I actually do quite a lot of things as a
5:32 normal data analyst. I don't really do
5:34 reporting and I don't build operational
5:37 dashboards. I work on quite big
5:39 open-ended questions which require a lot
5:41 more time and dedication and depth in
5:43 order to answer them. The way that I
5:45 would describe what I spend a lot of
5:48 time doing is I obsess about the why.
5:50 Why is our user base growing? Why is our
5:52 user base turnurning? Why have we been
5:54 able to sell more products to this type
5:56 of customer? And what can we do about
5:59 all this? Most of my insights work comes
6:01 from quarterly planning sessions when we
6:03 meet with business stakeholders. My
6:05 stakeholders are growth, marketing,
6:07 product, and some operations. We would
6:09 then go through what are the key product
6:11 launches, what kind of growth are we
6:12 trying to drive, and where do we feel
6:15 like we need data support? And this is
6:16 where the projects that I'm working on
6:19 get identified. Because my projects tend
6:21 to be quite big and open-ended, it's up
6:23 to me to decide how I approach it. It's
6:25 really important to have hypothesis or
6:27 clear questions in order to guide that
6:30 analysis so you don't get lost in all of
6:31 the different avenues that you could
6:33 take. After that I basically live in SQL
6:36 and Python and go deep deep deep deep
6:39 into the data. Every two weeks I try to
6:40 share something with my stakeholders.
6:42 Sharing insights is important because
6:44 that sparks discussion conversation and
6:46 that can help me guide formulate more
6:48 hypothesis off the back of the
6:50 discussions that I have. An output of
6:52 most of my work is some sort of slide
6:55 deck and I have produced so many in my
6:58 lifetime and I can truly say that I
7:00 spend equally as much time on the coding
7:03 as I do on creating a story through
7:05 slides and data visualizations. I think
7:07 a lot of analysts underestimate the time
7:09 that it takes to actually pull the
7:10 insights and turn it into something
7:12 meaningful that your stakeholders will
7:14 understand. And if you would like to
7:16 hear more about how to structure a good
7:18 PowerPoint presentation with your
7:20 analysis, let me know. It's definitely
7:21 something that I'm quite good at and
7:23 something that I'm passionate about so I
7:25 can definitely share in a future video.
7:28 I also never just send a slide deck in
7:30 an email or a Slack message. I always
7:32 schedule a presentation cuz that creates
7:35 an opportunity to have a discussion to
7:37 ask questions but also to make sure that
7:39 everyone is taking away what you want
7:40 them to take away because ultimately
7:42 it's your analysis. you know what you
7:44 have found. You want everyone to be
7:46 aligned on that story. It can also help
7:48 you get feedback on where your
7:50 interpretation isn't quite right. Your
7:52 stakeholders will always know more about
7:54 the product or the area and what is
7:56 happening which might help explain why
7:58 you're seeing your data in the way that
8:00 you are seeing it. I hope this brings it
8:02 to life a little bit. Yes, I quote a lot
8:04 and I live in SQL and sometimes Python.
8:06 However, I spend so much time
8:08 visualizing the data, thinking about the
8:10 why, and also presenting and
8:12 communicating with my stakeholders. So
8:14 those soft skills are equally if not
8:16 more important than the technical
8:18 skills. I think a more insights focused
8:20 role is a really good way to transition
8:22 from analytics towards data science
8:24 because in order to pick up those more
8:27 nuanced user behaviors, you can apply
8:29 regression, clustering or other machine
8:31 learning techniques to spot that. So you
8:33 while you will not be expected to build
8:36 productionready data science models, you
8:38 will get the opportunity to use real
8:41 world data on a real world problem that
8:43 your business is trying to solve and
8:46 that's an extremely valuable addition to
8:51 your CV as an aspiring data scientist.
8:53 Let's talk about compensation. Bear in
8:56 mind that I live in London in the UK and
8:58 this is based on my research as well as
9:00 my experience when recruiting for data
9:02 analytics roles last year as well as
9:04 this year. So for a graduate job or an
9:06 entry-level job, I would expect the pay
9:10 to be between £25,000 a year to £35,000
9:14 a year. For a mid-level with maybe 2 3
9:16 years of experience, I would expect the
9:19 pay to be between 40,000 to £50,000 a
9:21 year. For a senior that's talking maybe
9:24 four to six years of experience 50 to75,000
9:26 to75,000
9:28 and for a lead data analyst with maybe
9:32 five plus years of experience 75 to 95k
9:34 a year goes without saying that these
9:38 are very industry sector and company
9:41 dependent. If you work in public sector
9:43 such as NHS or government jobs or maybe
9:45 some of the media companies you will
9:47 probably get the lower end of the
9:49 bracket. If you work in private sector
9:53 such as banking, insurance, consulting,
9:55 obviously tech, you will get on the
9:57 higher end or some companies you know
9:59 like the big tech, they will pay above
10:01 the bracket. These are also based on you
10:03 being an individual contributor rather
10:04 than a team manager. With team
10:06 management comes more responsibility,
10:09 different type of skill set and you do
10:10 get paid more when you manage a team.
10:12 You would have also noticed that I
10:14 didn't put a six figure salary and that
10:16 is because I actually do believe that
10:18 it's quite hard to hit six figures in
10:21 data analytics in the UK and in London.
10:23 That is not to say that those roles
10:24 don't exist, but I think you will have
10:26 to be quite experienced maybe on the
10:28 lead/staff level and you will be
10:31 required to know some data science, data
10:34 engineering in order to land a six
10:35 figure salary. I generally found that
10:38 product analytics roles tend to be
10:39 higher paid, but I think it's because
10:41 you have to be quite a well-rounded
10:44 business orientated data analyst and
10:46 some of them do require a little bit of
10:48 data science experimentation knowledge,
10:49 which makes sense why they would pay a
10:51 bit more. But that is not to say that
10:54 there isn't growth or opportunity. I
10:55 think there's a very clear progression
10:58 if you continue to upskill and keep
11:00 yourself up to date with the latest
11:03 developments in the field.
11:05 I'm sure some of us have thought that AI
11:08 is an end of many careers as well as
11:11 data analytics. It is very true that we
11:13 definitely don't need as many data
11:14 analysts as before. We also don't need
11:17 as many analysts pulling basic queries.
11:19 I can already see some of my product
11:21 stakeholders writing code with AI rather
11:22 than waiting for the data team to come
11:25 back to them. But the reality is real
11:28 world data is always messy, sometimes
11:31 incomplete, and always needs context.
11:33 You need to know what data source to
11:35 trust, what is out ofd, and how it
11:37 connects to what's happening in the
11:39 business. All of that needs a human.
11:41 Most companies are also not startups.
11:44 They're very old and big and their data
11:46 systems are quite complicated and some
11:48 of them are regulated and it's going to
11:50 take quite a while for them to rebuild
11:53 their data infrastructure and their data
11:55 sources to be accurate enough for anyone
11:57 to write a couple of questions and
11:59 expect that the answer that it's giving
12:03 is using correct data. And that's a risk
12:04 that a lot of businesses are unwilling
12:07 to take. Instead, what AI is doing is
12:09 making analysts, engineers, or data
12:11 scientists a lot more efficient in what
12:13 they do. Now, I don't spend hours on
12:16 troubleshooting or correcting syntax
12:17 which doesn't work. I can do that with
12:20 IGBT. I can also spin up an ML model in
12:22 minutes. However, what I put into that
12:25 model needs to be created by me. Using
12:27 correct metrics and making sure that I'm
12:29 looking at the right data is more
12:31 important than the model itself. And
12:33 more importantly, my value as a data
12:34 analyst isn't coding. It's knowing the
12:36 business inside out. It's knowing what
12:38 metrics really matter and spotting when
12:40 numbers are off because of a weird
12:42 product launch that we had two years
12:44 ago. AI is not going to know any of it.
12:45 So, no, I don't see data analytics
12:48 disappearing. I see it evolving. AI is
12:49 making my job more efficient so that I
12:52 spend less time troubleshooting and more
12:54 time crafting impactful business
12:57 stories. And by focusing on turning data
12:59 into strategy and becoming a thought
13:02 partner to the business, you are
13:04 futureproofing yourself as well. And on
13:07 that lovely note, the sun has set and I
13:09 will wrap up here. Thank you so much for
13:11 sticking till the end if you did. Feel
13:12 free to drop me questions in the
13:16 comments or subscribe and follow for
13:20 more of this. Bye. [Music]
Chrome Extension