Connecting APIs Without Coding Skills To Create Your Own Dashboards – Yigit Konur at BrightonSEO
Articles,  Blog

Connecting APIs Without Coding Skills To Create Your Own Dashboards – Yigit Konur at BrightonSEO


Hello everybody, first of all welcome to
joining the session. There are 8 parallel session and all of all of them
are amazing so it’s a pleasure to see you. So today I’m going to talk about
connecting the API’s, without the coding skills actually. In my background
I was a dropout student in computer science, that’s why I have semi technical
skills, that’s the reason why I’m focusing on the stuff like this and
then helping people to use the technical stuff in better way without
having a CS degree. So let’s start with a question… Here is Istanbul my hometown, I am just living here but my career is getting
changed today, we are going to open an office in London and I am traveling here a lot. First of all, let me introduce myself, my name is Yigit Konur, I’m 26 years old
and a technical marketer and doing SEO since 2006/7 and I was 17 years old
and pretty younger than today. I also traveling a lot because of
speaking at several different conferences around the world. I also
just marry and I’m not traveling alone anymore so she’s here. I’m also one
of the judges of EU Search Awards and the United States Search Awards as
well and then I’m going to the conference’s and blogging over there and
then going to the after parties and drinking bourbon and then the most
favourite part of my career is actually our agency, Zeo Agency. I’m the founder
over there and we came here with eight different people and here is a picture
of our amazing team and then we also get a lot of the last Middle East North
Africa Search Awards and recognised as the largest SEO agency of the region.
Yeah the ads are over so let’s talk about the stuff we are going to see
today. And the last thing, we have offices in Istanbul and the capital city of
Turkey Ankara and in a few weeks we will also open the London office so here
is our amazing team and there was a video but it’s just not working for some
reason. So yeah let’s get working. So here is a complex situation, the API’s dashboard
JSON files, xml files… so I will simplify everything in just one story so here is
that story we have faced before. So this story is related with dashboards and if
you have dashboards you will not fail in the situations like this. And we were
working with one of the largest retailer of Europe and based in Turkey and for
sure I cannot share the name of the client. And then a SEO was one of the top
priorities for this client and they are paying attention to the organic traffic
and they are even tracking every stuff hourly organic traffic charges, etc.
And their team were really professional and we were doing great at this campaign,
all KPIs are satisfied and then they love us and then one day we get a call
from CMO and then, you know, in agency business and if one CXO the c-level
calls you for a reason it means that it’s not a great use for
you. So after than that we have seen that
there is a problem with the organic traffic so we need to fix this stuff and
it’s like the 14 percentage of the drop on the organic traffic’s and they are
still trying to investigate the problem. So at this time they did not have any
dashboards and the creative situation, just the web analytics tool and their
web analytics tool is not powerful enough to give exact insights to them so
they need to understand all the stuff by checking lots of different metrics. So
what we did in here is understanding the problem firstly. And
let’s trace the problem together, there is a huge drop in organic traffic’s like
40 percentage is a big problem, millions of traffic and 40
percentage of the drop means that millions of dollars actually. So when we
check the data actually in the page specific data there is no evidence on
the drop on the page types like the products and the categories and there
is not exact case on this situation. Then we talk with the IT team because
most of the situations like this, the instant drops, IT can be guilty so we
talked with them and then they are not accepting any responsibility and they
say that “we changed nothing and it’s not related with us.” Like usual IT. So after
them, we think that the possible reasons, maybe we’re penalised for some
manual algorithm penalty and we check the Search Console and see that there is
not a problem. OK and maybe it might be an algorithmic penalty and the second
one is the back-end team maybe create some problems and the 404 problems and
maybe the server error problems but there is not something like this. And we
also talked with the front-end team. Are there any content is getting removed
because you change your framework or some kind of a major
stuff. Nothing. And the last one is the data center actually and we talk with them because in some cases we have seen that they are blocking the Google IPs,
that’s the reason maybe the huge organic drop in here and they say that it’s not
related with us. So there is still infinite more reasons to review and have
a look but there’s a problem in here because the clients don’t know their
data. They have expensive web analytics solutions they think that they know
their numbers but they have no clue about what’s going on the website so as
an agency we have to help them and we need to help them.
So that’s why we offer to creating new dashboards to this client so after done
that we continue to do trace the problem so and then saying that we are always
analysing the data by URL perspective and understanding whether is the
category or the page or something like this and after then that we think that
maybe we can analyse a situation by breadcrumbs and actually the main structure of the website and analysing that and then understanding
category and the product traffics, like this example. And they have a home page
and then fashion and then clothes, maybe the men’s clothes. And then we want to
see that, the product and the category traffic separately and after done that
maybe we can understand that the problem is related with specific categories,
product pages maybe, so we have make some magic in here and then that get some
numbers. And the delta means that the difference between the last week, last
month, last year. And then what we see in here, the product traffic is totally like
the 20k but the difference is not that much big and the last month and the last
year as well but the category traffic is so high actually for some categories
and there is nothing related with seasonality actually when we are talking
about fashion, it might be related with a seasonality as well if you want to check
the Search Console it’s just coming from two days back so it’s also another
problem and also there is a great presentation today after than me. And
then what we need to do and after done that we check the situation and asked us
a question: maybe it’s the link juice problem, internal linking structure or
the category problem and the technical problem related with that. After then
asking lots of question and we finally ask that: are there robots.txt problems or
any no index or nofollow and we crawl the site with the Screaming Frog
and trying to extract all the metas related with the indexing the pages but
after that, the hours of the research and hopelessness, we have think that why not to check the Google Tag Manager. And actually, when we checked the Tag Manager, we have seen that somebody, and we still don’t know
who is it and IT takes responsibility over that, injecting some no index and no
follow text all of the specific categories and the product pages, which
means that if you are just doing the source on the Chrome you cannot see
this injected code because it’s just rendered by the JavaScript but the
Google is rendering JavaScript these days and then they are not indexing
anymore but they are the top product pages and top category pages and then
after then losing their index their traffic is getting dropped and it’s some
kind of a problem inside of the organisation I think. And somebody has a
smart idea to broke all the stuff and it works well but we fix the problem in
just few hours later. But what we learned over there is understanding data
from all aspect is really important and this case helps us to creating
dashboards for all of our clients. Also you should create your dashboards
as well because you have lots of reasons and understanding big picture helps you
a lot to create a new strategy and understand metrics better. So you can
also save hours instead of going to the one tool to the another one and it’s
time wasting and you can’t even set alerts to monitor all the organic
traffic. If something is getting dropped you can get a notification. I
will talk about it one by one but most of the people think that dashboards are
boring or just they are making the things for the executives and they want
to see the cool dashboards and the cool graphics like this but we are
approaching from a different perspective because it’s a need for us. And then we
trace the situation and then created our own dashboards and then I will show you
step by step how we did it. Ao the dashboard actually means that the
data plus visualisation and here we have a dashboard we wanted to create some dashboards and we need to start with the data. So when it comes to the data,
actually for the API’s we need to understand API and JSON. This is
so simple formats and from non-technical marketing perspective it looks like a
bit complex but I will talk with your language as well. So the API’s are just
like the communication style for the devices to the devices. Actually maybe
the web server of the deep crawl wants to connect with my Google sheets so they
need to communicate over the API. Even if I want to connect the deep crawl at
my mobile phone maybe I can use the internal API to do this stuff and it’s
like the mode of the devices and it’s like a language of the devices. Actually
I’m giving an example from our night life and if you are going to the bar and
then saying, “hey bartender, can you please give me one beer” it’s like a get
request on the API and you can understand this by the better way. If
you want to give something to the bartender like the tip, it’s like a post
request and generally we are using gets and post when it comes to API traffic so
you want to ask something or you want to put something to the server so we will
come back to this situation later but the bartender is here is the URL of the
server and the one beer is your requests details. And the other perspective is
the JSON as well. JSON is a grammar of this API communication generally and
there is also other grammars but the JSON is really easy to understand. As
you can see in here there is a user’s table and then the first name, last name and when they join to the company. The JSON is like
the structures text of the something and so you can just check the JSON in here
and here is another alternative is XML. XML is more complex but easy to read
and as you can see in here there is a person, it’s ID, the name, the start date,
salary, etc. It’s all written in here. So we also need to understand about the
JSON so it’s really easy to get all the details when you after them,
creating your first dashboard and there’s a great potential because most of the tools in our industry supporting the API so you
can connect all the stuff together, create your dashboard or create your own
technology by using them, it depends on you. I want to give an example from an
SEO monitor. Actually I want to create a dashboard to track my rankings and then
doing this stuff according to this. And then I’m going to say that, “hey bartender
I’m please give me one beer” and I’m saying that, “hey SEO manager please give
me all my rankings between last week and today” so it’s the computer language
is that the bartender is the URL and the request is here and then there is
authentication, it’s another story, but I’m asking the same situation in the API
language by just getting requests from SEO monitor. And
after than that SEO monitor answer me me and say “OK Yigit, you asked me
that all the related keywords between the specific dates, OK I’m listening to
you, on the name of the ID corner you are a listing for the position too and then
the URL is here and then search volume is here and lots of data in here. And
then I can use all the data according to my needs and actually in here I’m
just asking requests for Zeo’s account and then getting all the
keywords in here. So after done that, it’s so easy you are
just asking for the get request and it looks like easy to start but in here
it’s a little bit nerdy situation because most of the people are dealing
all the stuff by writing a script in Python or PHP, etc, and you can even learn
them online but it’s too much work and it’s not only in your major
actually. So maybe you don’t want to spend time with it. Actually at this time
you always have other options as well and as you can see in here and if you
don’t want to, if you couldn’t find the import situation so you can create
alternatives for you and we will not use any Python or the PHP scripts in
here but there is more visual solutions to this problem and not like this
solution but there is a great 2 tools I want to show to you.
The first one is Knime, the second one is rapidminer. These two tools are mainly
created for text mining. You are just doing this stuff in just visual ways on
their panel so it’s not like the terminal window and you can fix this easily and
what I say to you is if I want to ask a question to the SEO monitor and I want
to install the request module in Python and then need to write this code and
then link lots of stuff and it’s not that much easy. So instead of doing this
there is a much more easier way. It not looks like easier but I will
explain one by one, which is called Knime workflow. And here is a screenshot from my Knime and then what I’m doing in here is sending a get request to somewhere and
then getting this data filtered by an expert and row filter and filtering,
filtering, filtering, and creating tech clouds, colouring them and then creating a final product in here. And I will go one by one
but it’s really easy to use. You are just drag and dropping the stuff
and then creating your own workflow, it’s really easy OK.
And then as you can see in here ,here is my URL, here is my bartender URL and my
request and then sending the data and getting answer and then I’m filtering
also all the stuff by these filter modules and then I’m creating the final
charts, etc, and there are lots of works in here but it’s really easy when you go
to the YouTube and ask for the Knime tutorials and just go to the Knime TV, it
will help you a lot. And then you can connect all the stuff together. Another alternative is rapidminor. It’s a bit tougher than
Knime but you can still use it. So it’s a data perspective and let’s go to the
visualisation perspective and what we are doing in here is understanding,
making data more human friendly and as you know it’s a old verb: one picture
worth more than thousand words. And it’s also correct for the people because
humans are not talented as computers when it comes to process data so what to
do in here is understanding data visualisation. When you when I say to the
people to the data visualisation, the people always think something like this
but it’s not that much complex and I think that it’s useless actually. But the
data visualisation is that and I’ve just created a JSON file for all the BrightonSEO speakers and their previous years and the last year. Actually even
this is data visualisation because we are making this data much more easier to
understandable version for the people. So doing the stuff like this
is also data visualisation and as you can see in here, instead of using the
complex graphics even this kind of stuff is also data visualisation. And
behind of this visualisation there is still simple JSON files like this one and
as you can see some fancy versions in here but all of them are just coming
from the same input. So here’s what you can do to understand visualisation is
changing to data structures you are working on.
If you’re always looking for the marketing data you cannot understand the
pure idea behind of the data visualisation. That’s why I can
recommend to your openaddresses.io – it gives you different data to
analyse and then you can work on it. And also there is another great website, the
Timeline Story Teller – it’s also the Microsoft project and they have
created a timeline based framework for data visualisation. Just this table data
can convert it to this data and it’s kind of a different example but it might
be beneficial for you. And another great Microsoft product is
PowerBI and you can just use it and it’s free for up to one gigabyte of data
and you can make lots of visualisations in here and then it has lots of
connectors and you can just use this on your own workflow. And another great tool
is Supermetrics and it’s one of the best data visualisation solution
is Google’s Data Studio. A week ago they published new connectors for the Data
Studio. You might want to check it lots of tools in here you can just connect by
the Supermetrics.com. So finally it’s just all related with
making your raw data suitable for the input, for your visualisation tool and
it’s that much easy actually. And what we are doing in here is actually
like this: getting to data, shaping data and putting in the visualisation
tool and after than that you can connect different dashboards and then
make one big picture for each client. So another tool is dataviz.tools. There
is a great website and you can just check the stacks and then find the new
tools. And the final thing what I’m going to show today is a clip folio and they
have an all-in-one solution in here if you don’t want to spend too much time
but it’s less enjoyable because it’s managing all the stuff for you and it’s
not so customisable. And you can also create your own charts and then you can
manipulate the data and add some logic functions by using the clip folios clip
features. So what I can say to you is in creating dashboard is like this but you
just need to start from the one point and after then creating your one
dashboard for one client and it will come later for the other clients too. So
I don’t have enough time there’s a quick recap but you can just reach all the
stuff my username, I will publish on the Twitter. Thank you for listening.

Leave a Reply

Your email address will not be published. Required fields are marked *