Today's contrasty transparency film are not designed for shooting in bright sun. There is no law, of course, specifying that you do not shoot to gay abandon in the brightest and contrastiest light you can find (I've done it on occasion!), but when you mix shadows and bright light something has to be given up, and most meters will strive to bring up detail in the shadows at the expense of highlights (burnout). CWA of an at the margin area (partial shade and sun) can give a better reading, but multispot metering around a mid-tone can avoid either blocks (featureless shadows) or blown highlights. My favourite shooting conditions are grey, wet, overcast days which accentuate the colour of, e.g. rainforests that I have specialised in.
Velvia, as with most transparency films, blows gracefully, but blocks shadows at the slightest provocation. You might want to try 100VS as the first stop before moving onto the less forgiving Velvia stable. Stephen Schoof, above, makes a valid point about bracketing your exposures. Even a third stop adjustment either way will be noticeable, but often 2/3 is too much; so if the camera allows it, you can switch metering steps from say 0.3 (third stops), 0.5 (half stops) or 1 stop increments. Velvia responds well to either 0.3 or 0.5 stops. One big advantage of a handheld meter is to take several readings of different parts of the scene (as mentioned above) to get an idea of the scene's contrast and "fit" to the film's dynamic range. As a tool for analysis, this is worth the expense of a meter alone. This multispot metering of a scene was a big selling point for the venerated OM4.
iPhone meters are popular. An APUG friend of mine was using an iPod with an efficient on-board App meter for 35mm pinhole shots recently; can't recall exactly, but he may also have used it for SFX200 (faux-IR) exposures on that same walk. I'd still specify getting your hands on a decent spot meter to get a grip on the basics. Understanding metering from the spot meter's perspective is much easier than being confounded (or angered!) by unexpected results from evaluative / multipattern / matrix meters.