Sup.
I built a website (https://pompmeister.nl) which is essentially a big map with gas stations. I'm trying to find ways that I can improve its score on Google's PageSpeed Insights. As far as I'm concerned I've done pretty much everything–from compressing to enabling HTTP2.
Is there anything else I can do, or am I "crippled" by the fact that I'm using a map?
Thanks!
Related
I am looking to launch a new project and wanted to keep the latency as low as possible as well as load times.
i have been doing some research into CDN's and it seems like SSD CDN's are a good direction to go down but I have read that RAM based CDN's are even quicker!
I cannot find anyone who has done this, do you know anyone or how to implement this?
Using RAM for a CDN might work for really small scale systems, but not for globe spanning CDN's. Because RAM is volatile and needs energy to maintain the data. And energy costs...
There are solutions on the market, like non-volatile PRAM/PCRAM or in development, like RRAM (resistive-RAM), but not in the dimensions suiteable for a CDN. From my point of view, the tech is not there, yet. The current market (2014) is using SSD tech to speed things up. So if you want to start a project, focus on SSD.
You might want to read about CDN tech used at Google, Akamai, Edgecast, Fastly just to name a few.
Closed. This question needs to be more focused. It is not currently accepting answers.
Want to improve this question? Update the question so it focuses on one problem only by editing this post.
Closed 9 years ago.
Improve this question
I would like to know what are the main tools in the market for analysing/implementing E-customer behaviour in a Web application.
I just know Google Analytics which tracks client-side activity but maybe there are many alternatives using client and server-side scripts.
I already posted this question on webmasters.stackexchange.com E-customer behaviour in a Web application, but it has been closed and cannot understand why!
There are a vast array of tools to analyse user behaviour on a website. Ecommerce or otherwise.
Google analytics has options like:
Ecommerce
Ecommerce custom variables
Goals & Funnels
Goal flows
etc. which are useful in understanding things like drop off points, conversion rates, typical customer path and other shopper metrics.
Other analytics packages useful for ecommerce / website behaviour:
Clicktale
Crazy egg
Getclicky
and more. Some of these have a live / spy feature that allows you to see what users are doing realtime.
And the best way is to actually watch a recording of your users behaviour complete with keystrokes and mouse clicks / movements.
User recording tools:
MouseStats
Ghostrec
Inspectlet
Mouseflow
Most of the above also do aggregated heatmaps and overlays to give you an insight into what users click most or what catches their eye etc.
Incremental improvement to your website:
A/B testing or multi-variate testing are all the rage now. With A/B testing be aware of local maximum and also avoid the common mistakes people make with testing. Google content expirements (formerly known as google website optimizer) now is part of Google Analytics and you can use this to do testing.
References / more reading for analysing and setting up ecommerce user behaviour:
Stalking Users
Driving online sales
Web Analytics Solution
First of all you need to choose a general propose Web Analytics Solution. Since you are an E-commerce you want to choose one that has good support for tracking E-commerce data.
Google Analytics is the obvious choice here not only because it's free but also because it's better documented and easier to implement.
depending on your size it may make sense to implement a more Enterprise level Ecommerce solution. You may want to take a look at Adobe Omniture and IBM CoreMetrics. They are much more expensive not only because of the licenses but on an implementation perspective. It may take months to implement one of these other tools and the cost of the implementation can be almost the same as the costs for the licenses. Still if you need more enterprise level analysis and integrations with other BI solutions it may be worth taking a look at these.
Note that Google Analytics also has a Premium Edition. This is a fairly new alternative and provide some extra features and early access to beta features.
Product Recommendation
Depending on your Ecommerce platform you might already have some kind of product recommendation or up-selling. You usually can improve these systems based on analytics data. There are just a few options on the market, and most companies doing this tend to develop their own recommendation engine.
If you're just getting started with it, it might be worth looking at LiftSuggest. I haven't tried it but they seem easy enough to implement and leverage Google Analytics data to improve cross-selling.
HeatMap
It's easy enough to implement and may provide some nice isights. I find them more distracting usually but every now and then you can make good use of them. The most common seen around are CrazyEgg and ClickTale.
Behavioral Targeting
This is a technique to customize your site based on a previous knowledge you have about the visitor in order to increase his conversion rate. Tools don't help you much here, since you have to customize your site and no tool can predict how to do that. One common approach is to create buckets depending on factors that you can infer. For example: Users with Internet Explorer might be less tech savvy and thus might be more interested in non-tech products. On the other side Linux users are probably on the technology field. So you can put users on buckets depending on which country they came from, which browser they're using or if they are logged in you can use the information they entered on their profile or based on previous purchases. One tool that helps you doing that is called BTBuckets.
A/B and Multivariate Testing
Google analytics has an A/B testing tool integrated with the tool. Another Good tool that provides both A/B and Multivariate testing are Unbounce, Optimizely and Webtrends Optimize.
Custom Solution
Everybody these days are developing custom solutions. If you still have money and time to spend on Web Analytics after you exausted the other options you can look into building your own. Collecting the data the way you want and analyzing the granular data. Here solutions range fro server side to client side collection, but for the analysis they are usually done with Hadoop or with a OLAP Business Intelligence Tool like Microstrategy.
What you're looking for is a called Customer Relationship Management software, or CRM. They vary greatly, so without an in-depth understanding of your exact needs it's impossible to recommend specific ones. Any good CRM will let you analyze your site visitors in various ways. For example, you can see if customers bought X, they often came back and bought Y one month later.
The difficult part is integration because these systems need information about orders and other user actions. If you're using an off-the-shelf e-commerce package, there are often CRM options readily available.
For a "lighter" system you can use Google Analytics or similar, since it lets you send tracking, conversion, and sales information from the browser. It's great for analyzing the overall success of the site and tracking user actions across pages, but less powerful for sales-specific reporting and analysis.
I recently "inherited" an IT company that has a creative wing that does websites. Business is very good, but it seems like our throughput is very low. It's taking 3-6 months per site for just basic sites (no ecommerce, etc). From what I can tell, the process goes like this:
designers design the site using Adobe products
once those mock-ups look good, we outsource the files to be "sliced"
once we get the artifacts back from the "slicers", the web coders plug the information into a Joomla site
the site goes live
My question: is this the correct approach? I don't know enough about CMSs to know if we're using them in the way they're intended to be used. If there is a better process that you know of, or a better CMS that is easier to plug the results of the slicing into, I would love to hear about it.
Thanks for any feedback you might provide.
Ughh. I hate the process you describe. Designers who don't understand much HTML do the design, and then this "slicing" process involves cramming stuff that was never designed to be HTML into HTML.
Unfortunately, at my agency, we've found it difficult to find folks who design from the get-go in HTML/CSS.
But to your question about time, consider this. I'm a sole web developer working for an agency. We spent 4 months rebuilding our own site (a pretty simple site). But 3 of those months were spent going back and forth with the designer making changes.
I didn't start work on the site until one month before launch. From that point all I had was Photoshop files. I recreated the look of those files in Drupal templates and did all the other development (including some medium-complexity javascript) in just one month.
During that month, the design of 40% of the site changed dramatically. If it weren't for that, I would have been done in two weeks.
And I'm just one guy. And I don't consider myself a fast coder AT ALL. Your folks are taking WAY too long.
The web programmers should be able to slice the files, at the end of the day, they are the ones programming it so they should know what they want/need more than anyone else, it will save money and time theoretically.
Wow, 3-6 months for the workflow you described sounds painful. That's a timeline better suited to a site that includes a high level of custom data handling functionality. However, beyond the considerable mismatch between the process you describe and the time it's taking, the scope of this question seems pretty wide for stackoverflow.
I'd start by editing your question to indicate the amount of time typically spent on each of these steps. Just based on what you list I can't possibly imagine that taking 3 months, let alone 6.
Ok, if "Business is very good" and you're spending "3-6 months per site" then the only conclusion is that these are very high design websites with good margins. So you don't sound like you're in the slice and dice $14.95 a month website business.
Given that the sites we build often have a primary stake holder who is subject to corporate design guidelines and marketing criteria followed by product positioning criteria it's not unusual for us to spend 3 months getting the design work done. We can go through 3 design iterations (all of which the clients pay for).
The only difference to your situation is that our designers are 'web designer' and either understand how to build a Photoshop file to be 'sliced' or work in conjunction with the people that will be doing the slicing.
So, the only thing I find odd is that you send out the work to be sliced.
Nothing wrong with the process but the resource arrangement seems not correct. If it's a design company, I would expect you should have your own designers who can slice those images, rather than outsourcing. This will improve your efficiency and speed.
As for the CMS, Joomla is a much complicated CMS which is suitable for larger websites that require extensive CMS features. For normal corporate websites, I would suggest to use WordPress. It's very lightweight, easy to use, easy to skin and easy to setup.
Closed. This question is opinion-based. It is not currently accepting answers.
Want to improve this question? Update the question so it can be answered with facts and citations by editing this post.
Closed 9 years ago.
Improve this question
Is there a comparative analysis available for Mint and Google Analytics which can help me decide which one to implement in my situation?
For every website I work on I only use Google Analytics. My main reasons:
free -- for unlimited sites and hits
no ads
I don't find it hard to use and my clients haven't found it hard either (most figure it out within minutes)
it has an API, so if you want more, you can add your own stats or even incorporate into the site
for ecommerce, the extra tracking for ecommerce is great
there are theories that it might even help your Google ranking (as they gain extra data about your site)
a lot of people use it, so often the JS is cached
easy to setup (I don't know about mint, but GA is a few lines of JS)
connects to adsense so the stats can be tracked together (I don't know if others support this)
... and more.
Mint is simpler, always up to date, and has a lot of plugins aimed at blogs and personal sites. Google Analytics is more complicated and takes a little time to understand, is updated daily, and has more features for larger sites such as marketing, e-commerce and such.
On my own sites I started with Mint, used both of them for a while, but now I'm pretty happy with just Google Analytics.
I've been using Google Analytics in an enterprise environment for a few years now. There are certainly limitations, in particular with merging to other datasets. The API has been invaluable to bridging this gap, and really encouraging stakeholders to ask better questions.
In my opinion, evaluate both and work out the strengths and weaknesses based on your business requirements.
Why have you opted for these two possible options in particular? There are a number of good packages out there, many of which are free and offer excellent services that may provide compelling advantages over the "default" choice, Google.
Have a look at this ReadWriteWeb post that compares 10 free analytics tools for starters.
I use Clicky on my site and find it to be an excellent alternative to GA.
If you are interested in open source server-side analytics you should check http://piwik.org/. I did not use it myself but I've heard good it's decent, comparable to Google Analytics. I myself use GA and I'm happy with it. Just get some additional testing like Kiss Metrics and you should be fine :)
I recommend GA so that as your site grows bigger, you are already on top of a fine analytics tool. Mint is fine for now, but you need to plan for the future, too.
Anybody had experience using either of the following;
MapQuest platform: http://platform.mapquest.com
OpenLayers: http://openlayers.org
Which of the above would you recommend, any feedback from use of the service?
Or do you think it is worth paying for a commercial license from either Virtual Earth or Google Maps, get more available documentation online and more users to get knowledge and help from with the API?
Cost is an obvious issue hence why we are looking at the free commercial options.
Many Thanks
When my company evaluated this, we eventually ended up buying Google Maps.
Mapquest, in my biased opinion, looks like crap.
Google Maps everyone is already familiar with, works well, and has a large enthusiast community
We could afford it
That said, Google Maps is expensive, starts at 10K. We never got into contract discussions for VE so I don't know what they charge.
Another thing you need to consider is if you're going to be putting hundreds or thousands of markers on the map. If so, no solution will work out of box for you - they'll all run too slow in javascript. In that case you'll need to add the markers in a custom tileset or tile overlay on the server. And I suspect there are a lot of latency issues that go hand in hand with that.
OpenLayers coupled with OpenStreetMap may be sufficient for your needs.
OpenLayers isn't directly comparable to the other mapping providers. It is only a front end script to display various maps, it doesn't include any maps of it's own. There are various free and commercial maps you can use with it (for example OpenStreetMap.org), and it is also capable of embeding things like Yahoo maps or Google maps and providing the same APIs for all of them.