Stocking the gear

Gareth    Planning    

Another big issue for any online retailer, which we’ve not really covered here so far, is the actual stock. In our case, we are selling branded gear, from companies that have distribution and warehousing organised in the UK. This does make life easier, as we’re not looking to source our own, unique, product from the far-east, or having anything made.

We still have to make decisions on what to stock, in what colour, etc. This is a daunting task for an established retailer with several seasons of experience behind them. For a new entrant, it’s a nightmare. If we over commit on stock, in an attempt to cover all bases, we could be left with lots of expensive gear that we have to ‘dump’ at near cost at the end of the season. If we under commit, we could end up annoying our customers and being unable to get any stock of ‘hot’ items.

Our particular market makes this somewhat easier, as non-clothing items tend to have a shelf life of 1-2 years, as opposed to a single season. Even the clothing side of our market is far less fashion led than normal high street stuff.

I spent some time recently at the Lowe Alpine showroom with our excellent rep Becky (if you’re reading this Becks, how about another 5% off for the generous plug here!). We spent the day going over the Lowe range for spring/summer 08, which is being pre-ordered by retailers now.

Here’s a few shots of the Lowe Alpine showroom…

In deciding what to stock, we’ve leaned heavily on the experience of the brand reps, used our own judgment on what we collectively like, and taken Gareth’s advice on what trends are hot, and what gear is good for each activity. For our first year (so 2 distinct seasons for outdoor, summer & winter), we’re really guessing.

In time, we will have sales and trend data from the shop(s) and website, to predict what lines will do well and help our forward ordering strategy along.

In particular, we really like gear from brands like Montane, who are all about lightweight, high performance gear, it’s these kind of judgments we’ve made, in terms of general trends. Another key factor for the outdoor market, is the sheer number of brands that produce product for the same end use. It’s tempting to stock gear from every major brand, a sort of shotgun approach. Whilst there are a few brands that sit on the ‘have to have’ list, in most cases we’ve tried to slim down the number of brands we stock, and pick 2 or 3 (as opposed to 6 or 7) per major product category. Hopefully we’ve got this right!

This approach helps us become more focused in terms of the product we offer Webtogs customers. Take waterproof jackets, as an example, we could stock £200 GORE-TEX or eVent jackets from 6 or 7 brands. In our minds, as a consumer, this is too much choice and could lead to a confusing product proposition. Better, we think, to offer 2 or 3 core offerings from well respected brands. This also has the all important benefit of reducing our exposure to stock, a good thing for a fledgling retailer!

Managing development

Gareth    Planning    

As I’ve mentioned in a previous post, we are developing the Webtogs site from scratch, using our own development team. Our 2 primary php/ajax/mysql developers are based in London, and our office is in Dorset, so it’s a remote relationship. As the world becomes a smaller place, more and more companies are out-sourcing development work to other countries or domestic locations. With this in mind, I thought it might be interesting to take a look at how we manage this.

We’ve all worked together for some years, so the concerns over trust and that kind of thing are long since dealt with, but remain core concerns for companies looking to establish new working relationships. What I’m talking about here, is communicating the design and features to a group of remote developers, who, bless them, have to then make it all work!

All the XHTML & CSS is done by developers in Dorset, whilst the core code (ajax, msql, etc) is written in London. Our work flow, at a simple level (and on a good day!), goes something like this:

We also use Basecamp, from 37 Signals. I think this is a brilliant bit of kit. I’m not going to bang on about it for long here, as it’s been well covered on many blogs in the past. We tend to use it for milestones, messages and as a file repository for specification type documents. Any large files we need to share are uploaded to our servers in London, saving on storage space with the Basecamp system (there’s a limit).

Here’s a screen shot of our To-Do page, early in the development of Webtogs…

In addition to this, we also use Skype, as, I suspect, does the rest of the world! The file sharing tool on Skype, in particular, is a great feature.

Any changes to css, etc can be made locally from Dorset. We have SVN running on our development server, so if there’s a catastrophic overwrite, we can roll back a version with little hassle. (hasn’t happened yet, mind!)

We try to document as much as we can, and we use change request graphics, with the date in the filename to convey more complex changes. We’ve found this so much easier, than trying to explain in text what’s needed.

A recent change request graphic, for the config palette on the large product view…

At the end of every week, we delete all the completed change request graphics, as in the past, it’s become a total mess, file wise, for larger projects.

Outdoor trade show – Friedrichshafen

Gareth    Planning    

Well, we had a big weekend planned recently, where Phil & I were off to Zurich and then Germany to attend Outdoor, a big trade show for outdoor brands.

As it turned out, it didn’t quite go to plan….

Design evolution

Gareth    Planning    

I thought it might be interesting to write a post on how various elements of the Webtogs site have changed over time. In my previous post, I talked a little about our fist design ideas, and what, ultimately, was wrong with them.

When we first sat down and designed Webtogs, we added lot’s of features and tried to anticipate everything a potential shopper would need or might like to use. As the design has matured over time, we’ve actually taken away far more than we’ve added. We keep repeating the same old mantra ‘simple is good’.

Here’s a look at version 2 of Webtogs (click the image to see a full size one)…

I’ve marked some elements on the image, that i’d like to talk about a little.

Search Header

We liked the idea of providing a big and obvious search box on each page, really drawing attention to itself. We are using predictive search on Webtogs, rather like Google, so we figured this would be a nice and intuitive way to find product. The trade off here, however, is the loss of screen space to show products, offers, etc. Fairly shortly after coding this design, we removed the big search header from every page, except the home and landing pages. As the design shows, we also included space for various ‘up-sell’ links. Opinion is divided on these. From a retailers perspective, they can increase sales, but many online shoppers find them annoying.

Side Bar Widgets

This was a simple php function, that allowed us to define and deploy custom ‘widgets’ to the sidebar, beneath the main navigation box. In the end, we felt the information provided here was just as easy to find in the footer, and placing them on the sidebar added unnecessary clutter to the page, taking focus away from the products.

4 Across Grid

We like the idea of product thumbnails being larger than average. We experimented with 4 across, 5 across and 6 across views. This mock-up shows 4 across, but in the end, we felt the trade off between products per pixel and decent image size was too much for 4 across.

Footer

Every site needs one, but on this design, it looks lost and bit pasted onto the page. We’re doing a new, better (!) one, as I write.

Version 3, below, incorporating these changes…

As you can see, we’ve made several changes to the layout. Hopefully the new design is more open and product focused. The thin search bar (blue one) at the top of the page was the hardest of all the elements to agree on. It’s a great feature to have search on every page, in the same place, our only desire moving forwards is to make search more prominent on high traffic pages (i.e. home page, landing pages, etc).

Here’s a look at the journey we took with the left side-bar…

Version 1 & 2 are significantly different. For one thing, version 2 sees the addition of global gender navigation (as opposed to a category based system in the first design). We’ve also added navigation via brand, something our marketing research showed as critically important. Version 3 is really only a cosmetic overhaul, with the aim of opening up the site some more, and finally moving away from grey headers and boxes.

It’s interesting how one change has a knock on effect to other elements on the page. For ages we’d had 2 grey boxes at the very top of the page, one for help and the other for account log-in. This worked ok when the left bar had a grey top, but looked totally out of place with the version 3 design.

Again, a look at the journey we took with the account bar on top (the help link for the final version will be in the footer)…

The keenly observant amongst you will notice that version 3 provides for a photo of the customer. (That handsome devil shown is actually me!). This is another thing we talked about for some time. I’ve not seen many pure e-commerce sites that do this, but I think it’s something that will happen more and more as all aspects of the net become more socially orientated. Our main thinking here is based on reviews and comments, rather than simple account stuff, as it’s nice to post a review/ comment on a product with more of an identity (we think!). It will be very interesting to see how many users bother to upload an avatar.

Getting out of the smoke

Gareth    General Outdoor    

Living in London is fairly frustrating as it is the furthest away in Britain you can be from mountains, which for an outdoor person is like living with half a lung. It does however have lots of jobs and I am sure I am not the first person who although dreaming of a life in the Lakes / Snowdonia / Cairngorms, has to put up with the wilds of SE London instead!

Aside from walks around the Capital, the main escape, as well as holiday time, is the weekend away. This usually entails getting up at some god forsaken hour to travel to another part of the country where they have the things I love best – hills. Now I have a sprog on order at the moment which is due for delivery fairly sharpish so I have been trying to cram in as much as I can so far this year. One of our recent jaunts was off to the Peak District.

The Peaks can elicit a bit of a snobby response from some folks who might consider the bumps they have not large enough. At only 3 1/2 hours up the motorway though, for us London dwellers, it’s one of the better alternatives (the best IMHO being the overnight train to Inverness / Fort William). Usually it’s just the lass and me but for our latest trip I pootled off with a group I go walking with down south. As per usual this year, the sun took an absence of leave (Incidentally if anyone does see him remind him it’s JULY) and we were faced with some slightly damp conditions. …

All was not lost however as we were at one of my favourite campsites – Upper booth in the Edale valley. For those of you who have not been, it is an oasis of remoteness located around the start of the Pennine way. Part of a working farm, facilities are very simple but the views are some of the best. Nothing beats watching the peaks fade to purple and the sun sneak off towards Kinder. Unless of course you happen to have a beer in your hand which of course I did

Friday saw us head off for a stroll up on the hills above Derwent reservoir. The weather
forecast had been solid rain for the weekend but although the sun stayed away it didn’t piss it down!

It was however a prelude to the next days trail, walking back along Edale Valley before heading up and over Hollins Cross, Mam Tor & Lords Seat. Even though the weather was dicky we had some fairly stunning views. Towards the end, we got a little shower but all ended up grand as we pootled off to the pub in Edale. I think Duncan tried to do his Borg impression at that stage using nothing but a fork and a petzl…….

Sunday saw us pack up but we still managed to get a wander on Stanton Moor before we headed back to the smoke, knackered but with our heads refreshed. A full weekends walking, and all courtesy of the Peak District. Can’t be bad!

First designs

Gareth    Planning    

Here’s a look at some of the first design work we did for webtogs. The principle was fairly clear from the outset, but looking back, it’s amazing how things have changed over time.

Critical to us was the need for simplicity and ease of navigation. The basic rule of successful ecommerce for us, is this: “If someone wants to buy a backpack, sell them a backpack, don’t try to sell them a tent”. It sounds simple, but many sites seem to get this wrong, and annoy potential customers.

Early, 3 column design…

I think, looking back, we spent too much time trying to make a bad design work. The box structure was very limiting, as we were constantly trying to fit lots of info into small boxes. Another test we used, was to print the designs out on A3 paper, put them on the wall and take a step back. Looking at this design in that way, showed the interface far more than the product. Not a good thing, as we can’t sell an interface.

The search box, top left, was also an issue. We wanted a big, bold and prominent search box, that just couldn’t be slotted into this design. Actually, looking at it now, there’s really a lot wrong with this!

The persistent shopping basket (top right), was an idea we’d all seen on play.com and really liked using. But, when we started trying to make it work with the product database, it just didn’t work. Play mainly sell items that have no attributes (like size, colour, etc). When you add variations into the mix, it becomes very difficult to make a small persistent palette work correctly. This is another lesson we learnt the hard way. In hindsight, we would have been far better starting with 50 actual products, and working the interface around them, as opposed to shoving products into a pre-designed interface.