Dominic Wellington

Cloud-native architecture doesn't have to mean hosted in the cloud. Especially at the large-enterprise end of the market, plenty of companies still operate their own data centers for perfectly good and sensible reasons — but they can still take advantage of cloud-native architecture concepts to do so. Of course you give up the benefit of effectively infinite scale and reach that running in a hyperscaler's network brings, but many use cases don't benefit from that flexibility. If your use case is stable and its performance and growth models are well understood, you can size your infrastructure to your expected workload and operate it yourself.

Benchmarking is useful, but be careful to define objectives up front. Performance benchmarks conducted by one vendor against their competitors' products are generally not considered very credible by practitioners, so they are not that useful as marketing tools. You may still want to conduct them for internal understanding, but you may struggle to generate realistic simulations of real-world usage. Benchmarks conducted by third-parties are better, although there is still suspicion when they are sponsored by one of the participating vendors. Different benchmark providers have widely varying levels of credibility, too, so make sure you're not wasting your budget. However they are run, benchmarks are expensive; it's worth planning carefully.

This is one of the hardest parts of competitive analysis. Win data is easy enough to get, not least because sales people are highly motivated to track that — but losses are a whole other story. Did the deal get shelved entirely, did the incumbent keep the business, or did it go to a competitor — and was it the competitor you thought it was, or a completely different one? However, it's important to do your best to understand losses, because without that information, you only have a partial picture of your market and your relative position within it. It's just going to require manual work and sampling rather than automated analysis across all of your deals.

Qualifying questions are key, and you can never ask too many — but you have to be subtle about it. Many companies have some sort of standardized list of discovery questions, and training firms will also offer templates, but while these can be used to guide your conversation, they should never be treated as a checklist to be ticked off in order with a customer. What you are always looking for in your qualifying conversations is a compelling event: why the deal has to happen now, and won't be postponed to next quarter or next year or indefinitely. Without that sense of urgency, your deal will always be at risk.

One of the biggest issues with data quality is that the people who have the problem are not the people who can fix it, and the people who could fix it have no direct stake in the outcome. You don't need to assume malice or laziness: all else being equal, people will do the right thing most of the time. The problem is that other activities they have more direct ownership in will take priority, so ensuring data quality gets pushed down the to-do list — or even entirely off it. The solution is the same as for all these issues at the intersection of tech and culture: make it easy to do the right thing. Make data quality a built-in aspect of data capture, not a laborious afterthought.

SEO is a hard discipline, not least because the landscape is changing constantly. Rather than try to do your own inevitably crude analysis, work with whoever runs outbound SEO for your company. On top of specific expertise, they will already have custom tooling in place to monitor keywords related to your own product, so it will be a simple matter to work with them to get competitive monitoring added to their list. SEO specialists will also probably have all sorts of additional ideas that you might not have thought of from a competitive analysis perspective. Managed right, this can be a very valuable partnership, making your messaging far more effective — because you know exactly what your competitors are saying and how to counter them.