The personal experience of developing a travel website from scratch: mistakes, thoughts and ideas
August 23, 2017Comments (2)
We continue to publish translations (from Russian) of Victor Pashinsky’s cases. We are sure you will find a lot of useful things in this article and hope they help you avoid some mistakes at the start. For reference: Victor’s sites are targeted at the Russian-speaking audience.
Let’s here him.
There are no clear instructions in this article that could be applied “here and now”. But if you read in between the lines, you would find a lot of useful experience and ideas that would help you personally at the start.
Table of Contents
- Idea, available resources and previous experience. Finding and testing the niches
- The original idea of the resource
- Testing the idea and the first project
- Test results and conclusions
- First major resource and its development
- Different ways of monetization and their effectiveness
- Replacing the design and further developing the Montenegro travel guide website
- The second major travel guide website and its development
- Third travel guide website and tangible changes in design. Video upload
Idea, available resources and previous experience. Finding and testing the niches
Far back in 2014, I unexpectedly had quite a lot of free time of which I absolutely had nothing to do.
At that time, I had some initial skills on creation of simple sites – I could install CMS, theme and plug-ins. I also had a few dozen relatively popular stories. In general, they were the usual set of skills that you could acquire in a month or two if the desire is there.
The problem of users and its solution
Montenegro was my last visited country at that time. I had visited the country twice – the first time myself, and the second time with my bride. While preparing for the second trip, I noticed there were no good sites at that time containing information about Montenegro. There where no sites where I could quickly find descriptions of excursions or resorts, or big beautiful photos, etc. Even knowing the country, I could not provide visual image of it.
All that existed then – sites with repetitive 3×4 cm photos that could not be enlarged and a template text, which immediately shows that the person who wrote it has never been to the country or visited there 10 years ago. A set of articles on such sites was customized to the most popular keywords. You will be interested in something but the information is no longer there.
As a result, there was a usual situation for tourism: beaches in the country were crowded with people but there were nowhere to read about it. Search engine results were somewhat relevant, but the information found can barely be regarded as high quality. There was no information assembled in one place – you needed to search for information in hundreds of different sites, forums and resources.
Most sites actively promoted one of the other service, while information articles were only seen as a way to collect more traffic and not to help visitors.
The original idea of the resource
Looking back, I see that the original idea looks very naive. It is not viable in today’s world. Then, the plans were:
- To create a series of travel guide websites about different countries and places visited by me. One travel guide website for one site.
- To place them on the following subdomains to save money: montenegro.in-facts.info, bulgaria.in-facts.info.
- To promote the sites by placing links to each other.
- Monetization — only through contextual advertising somewhere in the sitebar. That is, separating content from monetization.
- Each site will have its own unique design (I was interested on seeing how different themes and plugins are arranged).
1. One travel guide website for one site
Now it is clear that of all the above-mentioned plans, only the first one can be considered reasonable. Indeed, to create one-size-fits-all travel guide websites dedicated to one topic is much more profitable than a blog with a set of posts about everything at once.
In my opinion, the era of travel blogs in the classical sense of this word has gone. Travel blogs were popular when there were just little content on the Internet and the search engine had to show at least something. Articles bearing such headlines as “look how I travel beautifully” occupied top search result positions only because there were no other articles.
But now, to get good traffic, it’s important to have structured and coherent information that gives ready-made recipes: “It’s better to lodge here, eat here, you can take shuttle bus No.1 for €2 or a taxi for €10 “. This is confirmed by the opinion and behavior of bloggers – they try to write the most detailed and accurate articles with a bunch of life hacks and useful tips.
Now let’s take a quick look at why the other four plans were wrong.
2. Subdomains and pennyworth savings
It would be much more reasonable to place each site on a separate domain. This would reduce the risks of losing traffic due to search engine pessimism. It would allow me to sell one of the sites if necessary and help determine what to place on the main domain. The choice of this domain itself was not a very effective one. It would have better to take something like travel-guide.info, etc.
3. Links to each other
Links from own resources do not help in promoting sites, and the influence of links is greatly exaggerated. In the niche of travel guide websites, I generally do not see any relationship between the number of links and the positions in search engine results pages. My new sites get to top positions in search results pages only in about 3-4 months after launching.
4. Monetization by advertising – not for tourism and it is very unreliable
It is extremely difficult to monetize a travel site by advertising only. To earn at least something, you would have to do the advertising too intrusive, otherwise you would earn $2-3 from 5000 visitors per day. Besides, you would completely depend on one or two ad networks, which does not fit in with my vision of a sustainable project.
5. A unique design does not promote branding and is very expensive to maintain
My resources now are just under 3 years old. And now there are people who once read the resource about Montenegro, and now they use travel guide websites for Bulgaria or Georgia. That is, by designing your sites in a uniform style, you may end up transferring part of the trust from the previous project to the next. If the designs were different, this effect wouldn’t have been there.
It also turns out that the sites tend to evolve over time. You may need new features, you may want to remake and update something. With 5 sites with different designs, you will spend 5 times more time on development than when the design was the same for all sites.
Testing the idea and the first project
Instead of taking up a huge work on a Montenegro travel guide website with hot eyes, I decided to test the idea on something less. In just a few weeks, 10 articles about the New World were written, a basic design was made and a domain was registered.
I suggest you take a look at the randomly saved screenshots and discover the main mistakes on them.
The website’s home page
Almost everything here is wrong.
Firstly, why would someone place cards on the home page of a site if the person has no plans to regularly update it with content? It’s better to place the general article for main keywords.
Secondly, there are social network buttons. In my experience from the sides, such buttons are practically ineffective. In fact, people very rarely use social networks on information sites — people prefer to send the appropriate link to friends. For social networks to be shared a little, they should be actively stimulated.
Thirdly, the beautiful photo above plays no big role. But at the same time, it attracts attention and is remembered most of all. As a result, we are no longer talking about branding since a completely different photo will be placed in another site.
Page containing a separate article
There are also quite a lot of mistakes here, which I would like to point out.
In the New World, Internet connection was very slow. Therefore, the suggestion here is that there should be a very extremely simplified version of the site without pictures, unnecessary CSS and JS. It sounds reasonable, but in fact I had to support 2 separate versions of the site and constantly address content duplication problems. All these were done without a special material benefit (there was no advertising on the simplified version).
A large headline picture in the article does not bring any benefit, but rather takes up a lot of space. The screenshot was made on a Full HD screen, and most of the visitors have ordinary laptops with a 1366×768 resolution, where the photo occupies the entire first screen. Some visitors simply left the site, not knowing how to scroll down.
It is preferable to make the headlines of articles longer by inserting into them several additional keywords. It turned out that the idea that an article with a headline exactly like the keyword will be ranked higher was misplaced. On the contrary, the attractiveness of such a snippet is lower in search results.
Content was placed in tabs. This is not a very good approach in terms of SEO and is not convenient for users (because they aren’t used to it).
On the right is a “standard” widget with a list of the latest articles. It’s not clear who needs such. Judging by the click map from Google Analytics, the widget was almost not used by anybody.
Test results and conclusions
Paradoxically, even in this form, 10 pages of the site quickly occupied good positions in search results and began to generate traffic. By the next summer, traffic already exceeded 100 people per day, and now 250-300. So, having 40-50 people viewing one article per day is to me a good result, but then the criteria were a little bit smaller, and the enthusiasm more.
I liked writing about travel, and so the experiment was seen as being successful. I started creating my first major travel project — Montenegro travel guide website.
First major resource and its development
To begin with, I highlighted the most interesting topics that could expand: resorts, prices, excursions, attractions, etc. Having created draft articles and grouped them by headlines, I received a simplified semantic kernel for 25-30 key articles.
The articles were written long and as detailed as possible. Many photos had our (my wife and I) faces — it’s necessary that people see that we really visited there and that we know what we are talking about (or maybe there were no other photos to upload?). Each article was divided into sub-headlines containing keywords.
If you look at the remaining screenshots of the very first version of the design, you would notice that it has become much more concise and of better quality. Here’s the home page as an example:
I would like to note the following errors:
- General design — too strict and serious. Considering that girls and women make up most of the audience, this kind of site appearance does not fit well with tourism and can scare them away.
- Having two columns (on the left and right) is entirely unnecessary. even with one side column. It’s already a bunch of useless links (even with one column) and gives a large visual load, which distracts the viewer from the main content.
- Slider — despite the beautiful appearance, there is no special benefit from the slider. Almost no one used them.
- Subscription form – collecting email addresses has no special meaning for travel guide websites of which there are no plans to regularly update them. There is simply nothing to send to the collected email addresses.
- Placing links to last entries instead of writing long articles – we have already dealt with this a little earlier.
Let’s consider a separate page of an article. In fact, the same mistakes as in the previous design of the New World site are repeated here – except perhaps the fact that a table of contents is added, which again was not used by anybody.
The site worked in this form approximately from January to October 2015 – that is, it managed to capture one summer season, evaluate traffic and check various ways of monetization.
Different ways of monetization and their effectiveness
I installed contextual advertising on the site when it started attracting about 1000 people a day. The results immediately saddened me. With unobtrusive placement of ad blocks, revenue from 1000 displays barely reached $0.3, which did not fetch me a dollar a day.
By experimenting on placement of ad blocks, I was able to increase this figure to about $0.7 with 1000 displays (from the editor: these indicators are usually much higher for non-Russian segments). Long articles and a large number of photos, which divert attention much from advertising, played a role in this.
I continued to develop the site. By closer to August-September, I registered in the Travelpayouts affiliate program and in several other affiliate programs. It just so happened that I literally immediately “dropped” several bookings, which more than blocked advertising revenue. From this moment, I started active development of these travel guide websites.
A Ukrainian version of the site, which quickly occupied top positions in search results pages but did not start bringing major income, appeared about in the same period.
You can find out all option how to make money from a travel blog:
Replacing the design and further developing the Montenegro travel guide website
Then it seemed to me that development of sites was being hampered most of all by a not-too-good design. I could not make the sites fast and neat. As a result, closer to October, the original design was changed and became quite similar to what you can see now.
To me, replacing the last entries on the home page with one detailed article was a key change, which secured better search positions in response to a keyword queries “Montenegro” (in Google) and “vacation in Montenegro” (in Yandex).
Individual blocks are gradually added or removed. I look at some of my colleagues’ ideas, do something myself, test something, and if I do not use a certain feature, I remove it.
If you want, you can take a look at a fairly typical article about Boka Bay. There were mistakes like having a big picture at the beginning, a huge number of the same type of illustrations and not too coherent narrative. But nevertheless, the article is quite working.
The site kept on developing, Russian-language articles reached almost 100 in number, and articles customized for specific affiliate programs appeared. Traffic also grew quite rapidly and by the summer, it exceeded 3000 visitors a day. Income also increased significantly — especially from affiliate programs.
Around May 2016, we suspended developing the site. We only sometimes updated old articles, updating information or taking into account changes in affiliate programs.
The second major travel guide website and its development
By August, the income from the Montenegro site has grown so much that I was able to go on a trip. So after all the experiments, I now created a Bulgaria travel guide website. In fact, little has changed — the articles have become a little shorter and laconic. They had less beautiful photos and common phrases, and more practical recommendations.
Bulgaria grew and developed a little faster than Montenegro, but it is contextual advertising that generated a significant part of the revenue — for 1000 shows, the revenue was 1.5 times higher than on the Montenegro site, and the affiliate programs are less efficient.
Now, this travel guide website is visited by about 2000-2500 people a day. There is a feeling that it is the most popular in the Russian segment on the Internet in its country.
Third travel guide website and tangible changes in design. Video upload
Georgia travel guide site, which was launched literally at the end of April, became the next notable round of development.
There were quite tangible changes in the design — a mini table of contents appeared, attractive ad blocks were introduced in the sitebar, a panel with advertising articles were added at the top and a beautifully designed authors’ section was also implemented. A photo at the very beginning, which was of no use, was removed. The number of photos in one article was reduced, but descriptions became more detailed.
Part of this feature was moved to other travel guide sites.
A short video appeared in many articles, which allows you to clearly learn about the place, resort or attractions. People indeed look at them. I think this is a positive factor for SEO, but for now I can’t say that with certainty.
I also like the authors’ block at the end of the article, which really stimulates you to share it in social networking sites and leads people to the article about independent vacation planning.
Interesting observation about gratitude. My e-wallet account numbers and links to my social media profiles were placed in the “Contact Us” page of all my sites.
On average, I receive about 3-4 Thank You messages in a week on the social media (links are placed only on the “Contact Us” page). These satisfied visitors tell stories of how incredibly useful the site turned out to be and how much it helped them. I receive more thanks in the comments section on the sites.
For almost three years the sites have existed (and several million people have visited), no one has ever transferred a penny to our accounts.
I think that’s all you need to know about people’s desire to support the resources that help them. I think it’s of no use counting on donations from users in the Russian-language community on the Internet – at least in the travel segment.